Environmental Management (2013) 52:1023–1039 DOI 10.1007/s00267-013-0149-y
Environmental Effectiveness of Swine Sewage Management: A Multicriteria AHP-Based Model for a Reliable Quick Assessment Marco Vizzari • Giuseppe Modica
Received: 12 February 2013 / Accepted: 8 August 2013 / Published online: 24 August 2013 Ó Springer Science+Business Media New York 2013
Abstract Environmental issues related to swine production are still a major concern for the general public and represent a key challenge for the swine industry. The environmental impact of higher livestock concentration is particularly significant where it coincides with weaker policy standards and poor manure management. Effective tools for environmental monitoring of the swine sewage management process become essential for verifying the environmental compatibility of farming facilities and for defining suitable policies aimed at increasing swine production sustainability. This research aims at the development and application of a model for a quick assessment of the environmental effectiveness of the pig farming sewage management process. In order to define the model, multicriteria techniques, and in particular, Saaty’s analytic hierarchy process, were used to develop an iterative process in which the various key factors influencing the process under investigation were analyzed. The model, named EASE (Environmental Assessment of Sewages management Effectiveness), was optimized and applied to the Lake Trasimeno basin (Umbria, Italy), an area of high natural, environmental and aesthetic value. In this context, inadequate disposal of pig sewage represents a potential source of very considerable pollution. The results have demonstrated how the multicriteria model can represent a very effective and adaptable tool also in those decision-
M. Vizzari (&) Dipartimento Uomo e Territorio, Universita` degli Studi di Perugia, Borgo XX Giugno 74, 06121 Perugia, Italy e-mail:
[email protected] G. Modica Dipartimento di Agraria, Universita` degli Studi Mediterranea di Reggio Calabria, Loc. Feo di Vito, 89122 Reggio Calabria, Italy
making processes aimed at the sustainable management of livestock production. Keywords Livestock sewage management Environmental assessment and modelling Multicriteria decision analysis (MCDA) Analytic hierarchy process (AHP) Decision support systems (DSS)
Introduction Agricultural intensification in the most suitable agricultural areas is one of the features associated with the more recent and significant landscape transformations that have occurred in Europe (Modica and others 2012). The environmental impact of high livestock concentration appears particularly significant where it coincides with weaker policy standards and poor manure management strategies (EEA 2007). Livestock sludge, in accordance with the best agricultural practices, can be used successfully as a resource and an organic fertilizer for crops, ensuring optimal disposal of such substances (Martinez and others 2009). In many cases, however, agronomic and environmental damages can result from the improper use of sludge, such as damage to the soil, degradation of the soil structure etc., because of the levels of certain cations (K?, Na?), salinization, alterations in soil pH, alteration of the soil microbial population, and accumulation of heavy metals. These processes have a tendency to lead to degradation of the agronomic potential of agricultural lands, and to the pollution of ground and surface water (CRPA 1993; Steinfeld and others 2010). The element causing the greatest number of management problems with regard to impact on near-surface and
123
1024
deep aquifers is the excess of nitrogen (N) (Teira-Esmatges and Flotats 2003; EEA 2005; EEA 2009; FAO 2009). This form of pollution, especially in relation to improper agronomic sewage utilization techniques, continues to represent a significant problem, especially in regions of intensive pig farming, in all the European Union Member States (APAT 2005; EEA 2009). Its related environmental costs are estimated at between €70 and €320 billion per year (Sutton and others 2011a, b). According to Eurostat data, the livestock units of pigs in EU-25 (the European Union as it included 25 member states until 2007) and EU-27, after a maximum peak at the end of the 1990s, showed a downward trend with a moderate diminution between 2007 and 2010 (-5.6 and -5 %, respectively). In contrast, a general stability of swine populations can be observed in EU-15 between 2000 and 2010, while in Italy, during the same period, a significant increase occurred (?7.8 %). The progressive concentration of production in large scale farms is confirmed by Eurostat data and, in Italy, by the National Swine Breeders’ Association, according to which 2 % of the farms house 84 % of the total animals farmed in Italy (ANAS 2005). In Umbria, the Italian region where the study area is located, the problems associated with pollution of near-surface and deep waters remain significant, especially in areas of high swine farming density where the highest pollution loads were observed, particularly in terms of N, BOD, and COD (Regional Council of Umbria and ARPA Umbria, 2008). In the light of the problems described above, effective tools for environmental monitoring of the swine sewage management process become essential for verifying the environmental compatibility of farming facilities and for defining suitable policies aimed at increasing the sustainability of pig production (EEA 2009; European Commission 2009; FAO 2009). Livestock farming performance and efficiency can be improved by adopting many different building solutions and farm management strategies (CRPA 1993; Hatfield and others 1998; Rotz 2004; Piccinini and Bonazzi 2005). To this end, many studies have been conducted with the aim of developing indicators and models supporting the efficiency evaluation and improvement of many agricultural processes related to production, energy and environmental issues. Frequently, an econometric type approach is adopted using both nonparametric techniques (such as data envelopment analysis, based on linear programming tools), and parametric techniques (such as the stochastic frontier analysis and stochastic frontier production) [for an extensive review of these techniques see Mendes and others (2013)]. Concerning the environmental efficiency of nutrient use, very common approaches are based on nutrient balances at the farm (Lanyon and Beegle 1989; Gysi and Schwaninger 2000; Oenema 2006) or regional
123
Environmental Management (2013) 52:1023–1039
level (Kesner and Meentemeyer 1989; Ju and DeAngelis 2010), and also for developing indicators to monitor the effects of agri-environmental policies (Brouwer 1998; Parris 1998; Yli-Viikari and others 2007); the advantages and disadvantages of these last approaches have been ¨ born and others 2003; Oenema and analyzed thoroughly (O others 2003). Considering that direct measures of N emissions in pig farms are expensive and time-consuming, several complex methods designed for nutrient environmental impact assessment are based on dynamic simulation models of N fluxes developed at the farm (Berthiaume and others 2005, 2007) and at landscape scale (Theobald and others 2005). Regarding the assessment of greenhouse gas (GHG) emissions from livestock production facilities, the most common approaches are mainly based on life cycle assessment procedures [for a comprehensive bibliographic review see Bellarby and others (2013)]. The above indicated approaches are generally very demanding in terms of input data requirements and time needed for set-up, calibration and validation. Moreover, there is still a clear need for adequate methods for analyzing data from different farms to demonstrate the variation in indicator values and to test the possibilities of reducing environmental impacts using farm-scale modeling (Halberg and others 2005). Furthermore, to better understand and manage an effective sustainable landscape evolution, it is fundamental to implement tools outlining potential future scenarios (Di Fazio and others 2011). In this regard, multicriteria techniques have already been used in decision support systems for the environmentally compatible management and planning of livestock production (Jain and others 1995; Gerber and others 2008; Marie and others 2009; Tichit and others 2011), for the assessment of the environmental effects of N-loads (Morari and others 2004) and for the evaluation of conventional versus alternative farming methods (Degre´ and others 2007). Multicriteria decision analysis (MCDA) is a multiple criteria comparison method supporting the decision-makers who are faced with numerous and conflicting criteria/alternatives with the aim of making an optimal decision. To achieve this objective, at least two critical issues must be faced (Tzeng and Huang 2011): defining the preference structure expressed by a decision-maker and identifying the correct weights of criteria/alternatives matching the decision-makers’ preferences. To this end, over the last 50 years, scholars have proposed several multicriteria methods and techniques dealing with theoretical and practical issues. Among these, the analytic hierarchy process (AHP) has been proposed to derive the relative criteria weights according to the appropriate hierarchical system (Saaty 1977, 1980). Considering its specific approach, AHP helps one to capture both qualitative and quantitative aspects of a decision and provides a
Environmental Management (2013) 52:1023–1039
powerful yet simple way of weighting selection criteria, consequently reducing bias in decision making (Ma and others 2005). Unlike previous studies that have used MCDA in the environmental analysis of livestock farming, the approach set out in this research has been aimed at the identification and analysis of those factors that, at farm level, mainly affect the environmental compatibility of swine sewage management. For this purpose, an innovative model [named EASE (Environmental Assessment of Sewages management Effectiveness)] has been developed and validated, enabling the integrated analysis of the said factors and the overall evaluation of sewage management process effectiveness. In consideration of the specific features and operational advantages of the AHP, a recognized robust and flexible multicriteria decision-making tool for dealing with complex decision problems as widely described in the following methods section, EASE has been implemented as an AHP-based MCDA model. Referring to the structure of this article, the ‘‘Materials and Method’’ section contains the scientific background of MCDA and AHP, and the conceptual and mathematical framework that supported the EASE model development. The same section includes the depiction of the study area where the model has been applied and validated for the first time, and the definition of specific scenario simulations in the same area. The results section, using tables and maps, shows clearly the numerical and graphical outcomes obtained from EASE application. The results are then summarized in the ‘‘Discussion and Conclusion’’ section where an analysis of the research findings is also provided, highlighting strengths and weaknesses of the model, and pointing out future directions for research to further validate and improve the EASE model here presented.
Materials and Methods MCDA and AHP: Background Concepts and Methods MCDA—or multicriteria decision making is an important field of operational research (OR) and management science. Scientific literature on MCDA methods is very extensive and goes back to the mid 1960s with the work of European scholars on OR (Roy 1968, 1996; Voogd 1983; Nijkamp and others 1990; Vincke 1992). Recently, a complete framework of the state of the art of the MCDA has been published (Figueira and others 2005); it reports on the most recent developments and open questions, including the use of MCDA for dealing with sustainability conflicts (social, economic, and environmental) at both micro and macroeconomic levels of analysis.
1025
MCDA can be synthesized as a complex and dynamic process supporting decision-making in combining the information from several criteria to form a single index of evaluation and, in turn, to make an optimal decision. Decision-making processes involve a series of steps: identifying the problems; constructing the preferences; evaluating the alternatives; determining the best alternatives (Keeney and Raiffa 1993). The specificity of MCDA lies in the determination of synthetic indicators based on several reference criteria, examined independently and interactively (Munier 2004; Roy 1996). Its purpose is to contribute toward the development of an iterative learning method feeding the evaluation process itself (Voogd 1983). MCDA methods can be divided into multiobjective decision-making (MODM) and multiattribute decisionmaking (MADM). When the decision space is continuous, MODM techniques, such as mathematical programming problems with multiple objective functions, are used. On the other hand, MADM deals with discrete decision spaces where the decision alternatives are predetermined. Depending on the mathematical algorithm adopted, MCDA methods can be classified as multiattribute utility theorybased methods, founded on functions allowing the ranking of all alternatives from best to worst, or outranking methods allowing the degree of dominance of one alternative over another to be defined. Within MCDA, a managerial and an engineering level can be distinguished (Duckstein and Opricovic 1980); while the managerial level defines the goals, and chooses the final ‘‘optimal’’ alternative, the engineering level defines alternatives and points out their multicriteria ranking. Normally, criteria can be quantitative and/or qualitative and have different measurement units. Over the last 20 years, multicriteria evaluation models have played an increasingly important role, due also to their integration into geographical information systems (GIS) decision-making tools (Pereira and Duckstein 1993; Malczewski 2006; Neri and others 2010) even supporting participatory landscape planning processes (Torquati and others 2011). Furthermore, as reported by Malczewski (2006), one of the most common procedures in spatial planning decision-making processes is still GIS-based MCDAs (GIS–MCDA), a process that combines geographical data (criteria) and decision-makers’ preferences. In these methods, the main rationale is that the two distinct areas of research, GIS and MCDA, can complement each other (Boroushaki and Malczewski 2010). In MCDA methods, well-known and widely used techniques relying on the comparative evaluation of alternatives/criteria are concordance–discordance analysis (Voogd 1983; Nijkamp and others 1990; Carver 1991) and AHP (Saaty 1977, 1980), both based on comparative functions. The AHP, in which the decisional problem is formulated through a hierarchical structure, is currently one of the
123
1026
Environmental Management (2013) 52:1023–1039
most popular methods for obtaining criteria weights in MCDA from a large amount of heterogeneous data (Saaty 1977, 1980; Saaty and Vargas 2012), also dealing with the uncertainty of unstructured or unquantifiable knowledge (Duke and Aull-Hyde 2002). In this respect, it is worth noting that in AHP the experience and knowledge of the people involved in a decision-making process are at least as valuable as the data they use (Vargas 1990). AHP is based on the principles of decomposition, comparative judgments and synthesis (Saaty 1977, 1980, 1994) applied in breaking down the decision-making model into a series of elements (objectives, attributes and alternatives) that must then be organized hierarchically (hierarchical structuring). In particular, decomposition involves constructing a hierarchy which puts the goal of the problem at the top and places criteria, sub-criteria and alternatives in descending order of the hierarchy (Saaty and Shang 2011). In other words, factors affecting the decision are organized in gradual steps, from the general, in the upper level of the hierarchy, to the particular, in the lower levels (Fusco Girard and De Toro 2007). Thus formulated, the problem is subdivided into several sub-problems for the comparison of the said elements by comparative judgments of criteria and sub-criteria with one another, so as to then place them in order with respect to a particular objective. The purpose of such hierarchical analysis is to find an order of alternatives in relation to the overall objective (Goal). AHP is based on iterative pairwise comparisons (PCs) between factors within a reciprocal matrix [PC matrix (PCM)] in which evaluations of the different factors are entered using the judgments from the experts. The comparative judgments envisaged are qualitative, exploiting the ability and the typical nature of human reasoning. Saaty’s fundamental scale of absolute numbers (Saaty and Shang 2011) is used to represent dominance with regard to a common attribute or condition by using ranges from 1 (indifference) to 9 (extreme importance, preference or
likelihood) (Saaty 1977, 1980). More precisely, the method involves five linguistic indicators that correspond to as many numerical values and, whenever necessary, they may also be conveniently used in intermediate judgments (Table 1). If necessary, decimals can be used to compare homogeneous elements whose comparison falls within 1 U. The need to increase the use of intermediate judgments increases as the number of factors considered rises. Considering that the weighting phase in the decisionmaking process is dynamically variable (Fusco Girard and Torre 2012), the PCMs of the EASE model have always been compiled by a working party of experts to minimize the subjectivity of the evaluations. In addition to aiding with the weighting phases, these PCMs enabled criteria, sub-criteria or alternatives being evaluated to be placed in hierarchical order and an effectiveness index to be calculated through, for example, a simple weighted linear normalization ranging between 0 and 1. AHP is widely applied by practitioners for its simplicity and because it allows relationships between factors (criteria and alternatives) to be established according to the decision-maker’s preferences expressed as ordinal language (judgments) and then converted into cardinal numbers. Indeed, most relevant criticism on the AHP method concerns the lack of mathematical foundation of the 1–9 scale used to convert ordinal judgments into cardinal numbers ¨ zcan and the resulting limitation caused by this structure (O and others 2011). Likewise, in situations in which there are a lot of alternatives and criteria the number of PCs can be so large as to represent a significant disadvantage of the AHP method. One of the most important advantages is that AHP, for each PCM implemented, allows for checking the inconsistencies in judgments provided by experts by means of a specific index [the consistency ratio (CR)] whose value should not exceed 0.1. If the CR is [0.1, it is necessary for experts to reconsider their judgments for a given PCM. As highlighted also by Saaty (2004), consistency is essential in human thinking because it enables us to order the world
Table 1 Saaty’s ratio fundamental scale of absolute values representing the intensities of judgments Definition
Intensity of relative importance
Equal importance
1
Slight importance of one over the other
3
Essential or strong importance
5
Demonstrated importance
7
Absolute importance
9
Intermediate values between two adjacent judgments
2, 4, 6, 8
1/9
1/7
1/5
1/3
1
Absolutely
Very strongly
Strongly
Moderately
Equally
Less important
123
3
5
7
9
Moderately
Strongly
Very strongly
Absolutely
More important
Environmental Management (2013) 52:1023–1039
according to dominance. The use of a PCM to obtain a ratio scale of measurement both for tangible and intangible factors is another recognized advantage of the AHP. In fact, the PCM effectively allows to overcome the human difficulty in simultaneously evaluating the importance of all the factors included in the evaluation. From these PCMs an absolute scale of relative values (importance) can be obtained through their principal eigenvectors and, if necessary, by normalizing them by dividing each value by the sum of all the values (Saaty 2006) or, as in our model, by the maximum value. As previously discussed, AHP is a hierarchical process. Thus, there is the need to combine the priorities of the alternatives derived under the different criteria belonging to the different hierarchy levels according to which the problem has been decomposed. To do this, an analytic network process (ANP), which is an extension of AHP and enables management of the interdependencies among criteria, can be implemented, developing a comparison matrix at each level of the AHP to compare pairs of (sub)criteria or pairs of criteria (Saaty 2004). This is also validated by the more general ANP interdependence feedback approach that involves the concept of dominance and raising the super-matrix to powers thus again using products and sums (Saaty 2006). Development of the EASE Model EASE is an AHP-based MCDA model aimed at making a quick assessment of the environmental effectiveness of the sewage management process. The preliminary phase of the model development has taken into account factors that may be considered determinants on the environmental effects associated with the management of pig sewage. In analyzing the entire farm sewage management process, attention has been focused on those factors with the greatest influence on the characteristics of the said substances, and on solutions for their disposal. Evidently, there are several farming and sewage management techniques that have an effect on these aspects (Sangiorgi and others 1986; Hatfield and others 1998; Bonazzi and others 2003). However, in light of the specific objectives of the study, the parameters selected were those that were most significant and easily determinable by direct surveys and short interviews with farmers. Considering the present is the first application of the EASE model, only the technical solutions most commonly adopted in pig sewage management in central Italy have been implemented in it. The different factors taken into consideration have been hierarchically structured in developing a model aimed at the calculation of a synthetic Environmental Effectiveness index (EEi), which reflects the environmental compatibility achieved at farm level in the said process. This index has been
1027
determined through the consideration of two main categories of factors: –
–
The quantities of sewage potentially produced by pig farms and the available farm land for its agronomic disposal; The technical solutions aimed at improving the effectiveness of the sewage management process.
In defining the core structure of the EASE model, two key indicators have been identified to quantify the aforementioned factors: the ‘‘Agronomic intensity’’ (Ai), defined by the yearly potential load of N per hectare compared with the maximum allowed load, and the ‘‘Technical effectiveness’’ (Te), capable of encompassing the major solutions aimed at reducing the impact of the sewage produced (Fig. 1). Indeed, a significant reduction of N emissions can be achieved operating on recovery buildings and on housing conditions (floor systems, manure removal strategy, diet composition, etc.), but also on sewage treatments and on final disposal solutions (Ds). For example, the adoption of partly slatted floors, the frequent manure removal, as well as flushing and separation of urine from faeces, enables NH3 emissions to be reduced (Philippe and others 2011). Within these management techniques, those that are conditioned by farming structural solutions (type of animal housing, sanitation techniques, primary treatment techniques) have been distinguished and combined within a structural factor (Sf). Likewise, those techniques that are associated with final sewage disposal have been identified and evaluated by means of a specific factor (Ds) characterized, within the model, by a significant influence on the overall Te. Indeed, suitable final disposals of the sewage (e.g., composting, anaerobic digestion, N removal) are key strategies in mitigating nutrients and GHG emissions (Bellarby and others 2013). As a consequence, a drastic increase of Te can be achieved, even in the face of a less than optimal Sf index. In addition to the conceptual plan, the model has been developed in mathematical terms with the aid of the aforementioned hierarchy analysis techniques. The EEi is calculated as follows: EEi ¼
Te Ai
ð1Þ
where Te represents the said technical effectiveness (between 0 and 1) and Ai is the said agronomic intensity which is determined as follows: Ai ¼
PNl PNlmax
ð2Þ
where PNl represents the potential N load expressed by the ratio between the quantities of N per hectare potentially
123
1028
Environmental Management (2013) 52:1023–1039
Fig. 1 Flow-chart of AHP-based hierarchical method for evaluating the environmental effectiveness of the sewage management process of pig farming
produced by the farm (determined in reference to CRPA 1993 considering the number and age of pigs farmed) and of the surface area used for agronomic disposal; PNlmax represents the allowed maximum N load, expressed in kg ha-1, defined in relation to the environmental vulnerability of the area (in this case set to 170 kg ha-1, in accordance with current Italian regulations). This easily determined indicator, even if related exclusively to the theoretical agronomic load of N, enables a summary evaluation of the potential pollution load of a farm in relation to the environmental conditions in which the livestock production is carried out. Equation 1 defines the basic configuration of the model considering that the environmental effectiveness of a farm depends mainly both
123
on the potential load of N and on the processes and techniques to reduce the polluting power of the sewage. The Te is calculated as follows: Te ¼
ðSf þ DsÞ 2
ð3Þ
where Sf and Ds are the said factors associated, respectively, with the structural and the disposal solutions. These parameters are calculated by means of an AHP evaluation. In more detail, the factor Ds is determined by means of a PCM by which an effectiveness level (ranging between 0 and 1) is calculated to be associated with the various disposal techniques considered (Table 2; Fig. 1). For those farms which adopt disposal techniques oriented toward an
Environmental Management (2013) 52:1023–1039
1029
Table 2 Effectiveness index of the Ds calculated using AHP Disposal solutions (Ds)
Criteria weights
Normalized effectiveness indices (0, 1)
Composting
0.44
1.00
Sent for purification
0.26
0.78
Separation and fertigation Burying of raw sewage
0.15 0.09
0.22 0.14
Surface spreading
0.05
0.09
We reported the normalized effectiveness indices utilized in the EASE model as well as the criteria weights (principal eigenvector of PCM)
external destination of the sewages (raw or partially treated) such as composting or external purification, the EEi index is set equal to Te considering that in these cases the Ai index becomes less representative in assessing their potential environmental effectiveness. The Sf is calculated by means of weighted linear combination, conventionally used in AHP for hierarchical aggregation of criteria: Sf ¼ ða AhÞ þ ðb WtÞ þ ðc StÞ
ð4Þ
where Ah is a factor associated with the animal housing typologies, Wt is a factor associated with the washing techniques, St is a factor associated with the sewage primary treatment techniques. The parameters a, b and c (criteria level), determined by means of a PCM, represent the weighting terms necessary for defining the relative importance of the three said factors (criteria weights Table 3). The latters (sub-criteria level) have also been calculated through three different PCMs, by means of which a hierarchical order for the techniques associated
with each Sf has been obtained (sub-criteria weights). The normalized eigenvector of the corresponding PCM, ranging from 0 to 1, expresses the level of effectiveness referring to each technique considered. All the PCMs have been developed in a spreadsheet, thanks to which it has been possible to implement and automate the entire evaluation model and proceed with the optimization and validation of the same through iterative remodeling of the attributions, and verification of the results with data collected in the field. This set-up has allowed to create a dynamic model that is susceptible to adaptation, expansion and remodeling operations in relation to the contexts studied and the subjects involved in the decision-making process. EASE Application and Validation, and Scenario Simulations The study area identified for a first application and general validation of the EASE model is the Lake Trasimeno basin, in Umbria (Italy) (Fig. 2). The basin, consisting of a natural basin and four artificially connected sub-basins (located southwest of the lake), is an area of high landscape, natural and environmental value. The area is also affected by numerous proposed sites of community importance and one zone of special protection established in implementing the ‘‘Habitat’’ 92/43/EEC and ‘‘Birds’’ 79/409/EEC Directives, aimed at the creation of the Natura 2000 European Network (the centerpiece of EU nature and biodiversity policy). The basin also contains the protected natural area of the same name, established by the Region of Umbria, and several areas of botanical interest and areas subject to landscape guidelines. According to the ‘‘Lake Trasimeno basin plan,’’ drafted by the River Tiber Basin Authority,
Table 3 Effectiveness indices of the Sfs (Ah, Wt, and St) calculated using the PCMs at criteria and sub-criteria level Sub-criteria
Sub-criteria weights
Normalized effectiveness indices (0, 1)
Criteria
Criteria weights
Perforated floor
0.66
1
Animal housing (Ah)
0.25
Washing technique (Wt)
0.25
Sewage treatment (St)
0.5
Partially perforated floor
0.26
0.4
Full floor
0.08
0.12
Dry
0.5
1
High-pressure washing
0.36
0.72
Low pressure washing
0.13
0.26
Nitrogen removal
0.417
1
Aerobic treatment
0.263
0.63 0.38
Anaerobic treatment
0.16
Separation and storage
0.097
0.23
Storage
0.062
0.15
We reported the normalized effectiveness indices utilized in the EASE model as well as the criteria and sub-criteria weights (principal eigenvector of PCMs)
123
1030
Environmental Management (2013) 52:1023–1039
Fig. 2 Geographical location of the study area
the entire basin has been declared a nitrate vulnerable zone of agricultural origin according to the provisions of the ‘‘Nitrates’’ Directive (91/676/EEC) of the European Community. The area houses several swine rearing farms producing sewage, the characteristics of which vary, depending on the farming and sewage treatment techniques adopted (Vizzari and others 2008). The agricultural census data confirm that the basin remains an area with relevant concentration of pig farms (ISTAT 2012). Considering the complete absence of sewage treatment plants, the obvious fate of livestock sewage is agricultural, being used to supplement or replace fertilizer components (Boggia and Pennacchi 1999). A survey form for collecting information useful for characterizing pig rearing facilities within the farms in the study area, with particular regard to the factors under consideration, has been prepared and performed for the purposes of application and validation of the model. The farms in question have been geolocated by a GPS device and the data have been imported in a GIS environment.
123
In order to test the model reliability, a scenario analysis has been developed. According to Tress and Tress (2003), our scenario analyses have focused on ‘‘what will happen if’’ instead of on generic ‘‘what will happen.’’ In more detail, we have designed five different scenarios to explore the effects of the improvement in technical solutions for sewage disposal for the pig farms in the study area. In defining the scenarios, we have taken into account some of the opinions of farmers, which arose from the interviews. Nevertheless, none of those is preliminarily taken as the most realistic scenario, but they have been defined to test the model potentiality to become an effective decision support tool. The scenarios can be summarized as follows: – – – – –
SC0: zero hypothesis, no change. SC1: Separation and fertigation in all farms with [1,000 animal unit equivalent (AUE). SC2: Building a composting plant in the north part of the study area. SC3: Building a composting plant in the central part of the study area. SC4: SC2 ? SC3.
Environmental Management (2013) 52:1023–1039
With reference to SC2–SC4, designed on the basis of building one or two composting plants in the study area, we assumed two baseline conditions for their best geolocation: a distance not [3 km from each pig farm and a preference for the locations near the larger farms (always in terms of AUEs).
Results Analysis of Farm Data During the site inspections, a total of 36 farms were registered with over 60 heads being reared. The various types of farm are characterized by different live weights for the animals being reared; thus, for a more objective comparison, the type of pig farm in question has been converted into Italian AUE units (equal to a conventional pig with a live weight of 80 kg), to which reference shall be made in the remainder of the study. A preliminary examination of the data clearly shows that the majority of the farms are specialized in the growing–finishing phase (25 farms in total) with 70 % of the livestock farmed. The eight fullcycle farms have *22 % of the livestock, while the remaining three breeding–gestation farms have the residual *8 % of the livestock. An analysis of the farming areas shows an uneven distribution around the basin area with a marked predominance to the west and a high concentration around the town of Castiglione del Lago (Fig. 3). Analysis, in terms of AUE, shows a clear predominance of medium-sized farms (20 farms from 500 to 1,500 AUE) corresponding to *39 % of the total heads of livestock farmed. Then, there are six farms with fewer than 500 AUE with only the *4 % of the total livestock and six medium-to-large-sized farms (between 1,500 and 3,000 AUE), corresponding to *23 % of the total livestock, and only four large farms (over 3,000 AUE) but with *34 % of the total livestock of the area under consideration. In summary, it can be observed that in the two largest size classes, there are ten farms rearing 57 % of the pigs present in the Trasimeno basin. From the farm data collected, an analysis can be performed based on the factors considered in the evaluation model. In relation to floor type, it may be observed that *48 % of the livestock (within 20 farms) are reared on mixed floor types (full flooring with a grille section) and *45 % on completely grilled floors (11 farms). The remainder of the livestock is reared on full flooring (*7 %). In general, the good incidence of grilled and mixed-type flooring indicates a good saving in terms of the amount of water used to wash the stalls, and, consequently, less dilution of the sewage, with obvious advantages in
1031
relation to the treatment and disposal of the same. The predominant use of high-pressure washing techniques (20 farms with *84 % of farmed livestock), and the much less frequently used dry cleaning techniques (four farms with 7 % of livestock), shows a certain degree of attention with regard to the effectiveness of these practices. Again in this case, these aspects can contribute toward a reduction in the quantity of sewage produced, with obvious environmental advantages. In relation to primary sewage treatment techniques, there is a high incidence of simple storage (71 % of the livestock) in open-air storage tanks, which is then, in the majority of cases, buried in the fields. Several farms have stated that they perform mechanical separation of the solid and liquid fractions and, in relation to the data collected, this is the practice in eight farms, raising 29 % of the equivalent head of livestock in the basin area. No other primary farm sewage treatment methods have been reported in the basin. From the data relating to the technical solutions adopted for sewage disposal, it may be noted that distribution over the fields and subsequent burying of sewage, following
Fig. 3 Geographical location of swine farms in the study area and their classification according to livestock dimension expressed as AUE
123
1032
Environmental Management (2013) 52:1023–1039
Table 4 Number of farms and animal units in relation to the PNl classes PNl classes (kg N ha-1 year-1)
Number of farms
1—up to 170
11
30
8,015
17
2—from 170 to 340
14
39
18,604
39
3—over 340 Total
11 36
31 100
21,445 48,064
44 100
% of total
Number of pigs (AUE)
%
storage, is the most frequently observed (23 farms with 57 % of the livestock). This is followed by fertigation, generally associated with those farms performing solid– iquid separation (eight farms with 29 % of the livestock). There are four farms in the basin area, with *13 % of livestock, where the technique of composting has been adopted. Although it involves significant initial investment, this solution allows for the excellent use of pig sewage through the transformation of sewage into compost, generally of high quality, which can be used in agriculture as a soil improver. In one production facility with 450 heads of pigs, it has been stated that liquid sewage is disposed of by
spreading it directly onto the fields. This inevitably results in severe environmental repercussions because of the high risk of run-off, with consequent pollution of water courses, emission of odors, etc. Evaluation of Environmental Effectiveness Using the information collected from the pig farms located in the basin, the intermediate indices envisaged in the EASE model (PNl and Te) and the EEi have been determined and analyzed thoroughly to evaluate the model’s consistency and reliability. As already specified, the PNl highlights the potential environmental pressures associated with the quantities of sewage produced by the farm. The values of the index have been sorted into three classes, defined by limits obtained from current Italian regulations (Table 4; Fig. 4). The limit value of 170 kg ha-1 year-1 represents the maximum N load applicable to vulnerable areas, while the limit value of 340 kg ha-1 year-1 represents the maximum load for other agricultural areas. From the examination of the data, it emerges that only 11 pig farms, responsible for rearing 17 % of the livestock,
Fig. 4 Geographical location of swine farms in the study area and their classification according to PNl and Te
123
Environmental Management (2013) 52:1023–1039
1033
Table 5 Numbers of farms and pigs according to the four Te classes Te classes
Te interval numbers
1—Excellent
0.75–1.00
2—Good
0.50–0.75
3—Moderate
0.25–0.50
4—Poor
0.00–0.25
Total
Number of farms 3
Number of pigs (AUE)
%
4,688
9.8
1
1,463
3.0
26
37,324
77.7
6
4,590
9.5
36
48,064
100
Table 6 Number of farms and pigs according to the four EEi classes EEi classes
EEi interval numbers
Number of farms
Number of pigs (AUE)
%
1—Excellent
0.75–1.00
3
4,688
9.75
2—Good
0.50–0.75
5
4,710
9.75
3—Moderate
0.25–0.50
9
7,186
15
4—Poor
0.00–0.25
Total
19
31,480
65.5
36
48,064
100
produce a PNl below the limit of 170 kg ha-1. On the other hand, 14 farms, rearing 39 % of the livestock, have a PNl [170 kg ha-1 but lower than 340 kg ha-1, while the remaining 11 farms, rearing the remaining 44 % of the livestock, are responsible for a potential load greater than the maximum allowed limit for non-vulnerable agricultural areas (340 kg ha-1). It is worth noting that in the study area, the swine farms can be considered a significant potential source of pollution also considering the closer proximity to the lake, particularly of those farms with the higher PNl. The Te index, by virtue of its structure that includes four key parameters evaluated by means of AHP, enables reliable quantification of the overall efficacy of the sewage management techniques adopted on the farm. The values obtained have been divided into four Te classes, characterized by having equal-sized intervals (Table 5; Fig. 4). The results highlight three farms classified in the ‘‘Excellent’’ class, with *10 % of swine livestock, while the ‘‘Good’’ class only contained 1 farm, with *1,500 AUE. The four farms of these two classes achieve very efficient sewage management because of the adoption of on-farm composting as the final destination of sewages. The majority of the farms show ‘‘Moderate’’ effectiveness in terms of technical management solutions (22 farms with 77 % of livestock). These performances can be explained mainly by considering the wide diffusion of the simple storage of the sewage that appears as a low effectiveness technique for reducing N content. The Te index of farms, in the moderate class, which adopt liquid–solid separation, is lowered by the use of a partially perforated floor that
inevitably determines a volume increase in sewage. The remaining six farms show poor technical management effectiveness mainly because of the adoption of full floor or partially perforated floor in association with low pressure watering. As described, the EEi is a complex index that summarizes the environmental effectiveness of pig sewage management at farm level. Also, in this case the values obtained have been divided into four efficiency classes, characterized by having equal sized intervals (Table 6; Fig. 5). The results show that 19 farms fall into the ‘‘Poor’’ effectiveness class that accounts for *65 % of the pigs reared in the basin area. Within the ‘‘Moderate’’ class, nine farms account for 15 % of the animals, while the ‘‘Good’’ class contains seven farms corresponding to 19 % of the total livestock. The ‘‘Excellent’’ class contains the same farms with higher technical effectiveness at the same time the other farms that adopt composting achieve good environmental effectiveness since the model, as specified, considers EEi equal to Te for this kind of disposal. The data indicate that only circa 20 % of the pig farming activity in the basin area is accounted for by eight farms that, in relation to the results from the model, manage to attain
Fig. 5 EEi of the swine farms in the Lake Trasimeno basin resulting from the application of the EASE model
123
1034 Table 7 Numbers of farms and pigs in the four EEi classes, using 340 kg ha-1 year-1 as NPlmax in the application of the EASE model
Environmental Management (2013) 52:1023–1039
EEi classes
EEi interval numbers
Number of farms
Number of pigs (AUE)
%
1—Excellent
0.75–1.00
13
12,348
25.7
2—Good
0.50–0.75
4
4,236
8.8
3—Moderate
0.25–0.50
12
20,163
41.95
4—Poor
0.00–0.25
7
11,318
23.55
36
48,064
100
Total
good and excellent environmental effectiveness. On the other hand, it may be observed that about 80 % of the AUEs are reared in 28 farms where the sewage management process is managed with overall moderate and poor environmental effectiveness. The environmental performances of pig farms, measured by the values of the EEi index, are strongly influenced by Ai values and, in particular, by the PNlmax level that, as indicated, represents a user-defined parameter that reflects the vulnerability of the area, taking into account the current environmental protection regulations. In a nonvulnerable area, considering the higher allowed N load, the same farms would attain better environmental results in terms of EEi (Table 7; Fig. 6). In order to better compare the results produced by the EASE application to the different scenarios foreseen, also in this case, we refer to the AUE occurring in the different EEi classes rather than the numbers of farms (Table 8; Fig. 7). As highlighted previously, the majority of pig farming in the study area is characterized by a high potential environmental impact (65 % of AUE fall in the EEi poor class). In this regard, it appears relevant to note how the adoption of separation and fertigation techniques in all farms with an AUE [1,000 (SC1) has no significant effect on reducing the environmental impact of pig farming in the area (54 % of AUEs remain in the poor class, while the total of AUEs falling in good and excellent classes remains the same). The two hypotheses regarding the building of a single composting plant (SC3 and SC4) have a very different effect on environmental impact. Indeed, SC3 is characterized by better environmental results because of the significant greater presence of pig farms in the central part of the study area compared with its north part. These effects, in particular, are more evident for the moderate and good classes. As expected, the best solution in terms of reducing the environmental impact of pig farming involves the building of two different composting plants for both the central and northern parts of the study area (SC4, Fig. 8). Following this hypothesis, the AUEs in the poor class fall from 65 to 20 %, with a valuable improvement in reducing the environmental impact of swine farming in the Lake Trasimeno basin.
123
Fig. 6 Comparison graph between the numbers of pigs (in AUE) within the EEi classes, resulting from the application of the EASE model, using the maximum nitrogen load allowed in vulnerable area (a) and in other areas (b)
Discussion and Conclusions Livestock production today is still a major contributor to the world’s environmental problems and its future plays an important role in the debate concerning the critical issue of environmental sustainability, as well as in the global food crisis (Steinfeld and others 2006, 2010; Sutton and others 2011a; Weiss and Leip 2012). Consequently, the environmental compatibility evaluation of livestock sectors, in various regions and climatic conditions, has important practical implications in the socioeconomic policies at the different administrative scales. Sewage management is a central topic in the environmental analysis of intensive livestock production systems and holistic approaches should be adopted in evaluating the different management scenarios. Traditionally, slurry and manure from farms have normally been used as organic fertilizers for crops, generally ensuring optimal disposal of such materials (Martinez and others 2009). However, as already specified, different agronomic and environmental damages can result from poor manure management strategies such as degradation of the soil structure, salinization, alterations in pH, alteration of the soil microbial population, and
Environmental Management (2013) 52:1023–1039
accumulation of heavy metals. These issues can become particularly relevant in areas of high livestock concentration characterized by weak environmental policies and poor sewage management strategies. Excess N is certainly not the only main risk factor associated with improper management of livestock sewage (CRPA 1993; Steinfeld and others 2010), but phosphorus can also give rise to significant environmental issues related to eutrophication of water bodies (EEA 2009). Reduction of these emissions requires a whole-farm approach based on different feed and management measures starting with an improved diet composition, to reduce animal excretions, until applying the appropriate amount of manure in a timely manner (Rotz 2004). Moreover, many building solutions can also contribute to significant emission reduction, as in the case of the adoption of slatted floors (Philippe and others 2011). Aside from the cited measures, storage and treatment options, and the disposal of sewages, are also very relevant concerns, particularly with respect to surface water and groundwater quality and to air quality (Sangiorgi and others 1986; Hatfield and others 1998). Regarding the study area, several different pig sewage management farming strategies have been observed. In some cases, the farmers consider sewage as a waste to be disposed of in as short a period of time as possible. This attitude is often the source of uncontrolled disposal of insufficiently stored sewage over an unsuitable timeframe and on a greatly reduced surface area. On the other hand, other farmers, aware of the possibilities and the environmental and financial advantages to be derived from the environmentally compatible use of livestock sewage, have shown great sensitivity and care with regard to the management practices concerning such materials. Some of the farms in the Trasimeno basin, for example, have already adopted the practice of composting, which allows for a drastic reduction in the environmental impact of intensive farming, producing materials of generally high quality that can be used in agriculture with excellent results for improving the physical and chemical characteristics of the soil. Besides these two opposite cases, there are several intermediate situations represented by farms making more or less rational use of sewage, even if over a somewhat reduced surface area and with often inadequate storage times. In these farms, the spreading of sewage can contribute to a reduction in mineral top-dressing, even though the tendency to overload can result in environmental risks, especially in areas characterized by greater environmental sensitivity. Thus, more careful management of the entire sewage production process can produce significant benefits for the quality and quantity of sewage used in the fields and, obviously, greater top-dressing efficacy. As a general rule, farm management must focus on improving N-use
1035
Fig. 7 Comparison graph between the numbers of pigs (in AUE) within the EEi classes, calculated for the different scenarios
Fig. 8 EEi of the swine farms in the Lake Trasimeno basin considering the hypotheses of SC4 (building two composting plants in the central and northern parts of the study area)
efficiency, retaining the emissions from the manure until it is incorporated into the soil, and applying a proper amount during the appropriate period to enhance crop uptake (Rotz
123
1036
Environmental Management (2013) 52:1023–1039
2004). Nevertheless, the more effective manure treatment technologies generally entail heavy investment and operating costs; thus, only a few techniques, such as composting systems, are successful options in Europe and North America (Martinez and others 2009). Results show that they also represent a very effective solution well suited to the Trasimeno ecological situation. Furthermore, these systems have a greater chance of success, especially in developing areas where the cost of chemical fertilizers represents an important cost factor, such as in the Asian countries (Burton and Martinez 2008), which now account for circa 60 % of the world’s pig population (FAO 2013). However, composting plants require considerable initial investment that may be difficult for an individual mediumsized livestock farm to meet. Therefore, in addition to suitable regulations for defining precise environmental protection restrictions and imperatives, there is the opportunity to include appropriate measures in the Rural Development Programs for funding composting facilities, possibly also organized as a consortium. This framework indicates how the building design, diet, and management measures (including sewage treatments and disposal options) that are effective in reducing emissions from livestock farming vary greatly in their type, applications, and related effectiveness. Considering this complexity, there is still a clear need for effective methods and farm-scale models for analyzing all the cited environmental measures and strategies, and evaluating alternative scenarios, to define appropriate actions to reduce the environmental impact at farm and at landscape level. As outlined in the introduction, the great majority of the existing modeling approaches are very demanding in terms of input data requirements and time needed for set-up, calibration, and validation and, frequently, are focused on few factors influencing the environmental impact of livestock farming. The EASE model, developed in the study, has shown excellent flexibility and ease of application to the various livestock facilities examined. Allowing a quick evaluation of the most relevant parameters characterizing the pig sewage management process can become a valuable tool to support the sustainability analysis at farm level, as well as on landscape and regional scales. The indices,
determined during model implementation for the various options by means of PCMs, reflect coherently the effectiveness of the specific techniques as derived from expert judgements. Moreover, the intermediate and final indices (Sf, Te, and EEi) allow a diversified and reliable quantification to be made of the most important environmental effectiveness factors associated with pig farming. As also demonstrated by scenario analysis, EASE can straightforwardly support a selection of the more environmentally effective solutions, including in terms of best locations for the composting plants. Unlike more complex approaches, such as the cited econometric and other mathematical simulation tools, the benefit of this model lies in its ability to perform an expeditious evaluation of farm environmental performance using easily determinable data and indicators, thus achieving clear operational advantages with regard to decisional processes. Thanks to the great adaptability and versatility guaranteed by the MCDA–AHP techniques, the model may prove a very useful tool for supporting decision-making processes aimed at the sustainable management of livestock production processes. The particular MCDA–AHP application performed in this research, because of its specific objectives and approach, can be considered innovative in the context of the existing literature about similar topics. The EASE model, given its inherent aim, includes the most environmentally relevant factors of the sewage management process. At this stage, it does not consider other cited, environmentally important (but not easily determinable) factors, such as the different diet compositions or the seasonal variations in the N load, as well as specific variables related to the energy efficiency of the farm. In order to improve EASE, these parameters should be integrated into the future versions of the model. In addition, a more comprehensive evaluation of the mitigation techniques linked to the housing conditions and to the entire manure management process (storage, treatment, and spreading phases) could advance the overall reliability of the results. As well as environmental factors, future improvements of EASE could integrate economic indicators to better identify the production/efficiency trade-offs and synergies also within scenario simulations. Future research directions
Table 8 Number of pigs (in AUE) within the EEi classes calculated for the different scenarios EEi classes
SC0 AUE
SC1 %
AUE
1—Excellent
4,688
10
2—Good
4,710
10
3,280
3—Moderate
7,186
15
12,811
4—Poor
31,480
65
25,855
Total
48,064
100
48,064
123
6,118
SC2 %
AUE
13
SC3 %
AUE
SC4 %
AUE
%
6,018
13
7,725
16
9,055
19
7
4,716
10
11,928
25
11,934
25
27
17,325
36
7,191
15
17,330
36
54
20,005
42
21,220
44
9,745
20
100
48,064
100
48,064
100
48,064
100
Environmental Management (2013) 52:1023–1039
should also encompass aspects linked to bio-security issues of slurry management, as well as nuisances (e.g., noise, odors, etc.) caused by transport and the spreading of slurry (Lopez-Ridaura and others 2009). Additional validations of EASE, also based on comparison of the results with other already validated indicators, will allow for further optimization and developments of the model, thereby improving its effectiveness in evaluating the environmental sustainability of the sewage management process, in view also of its application to other typologies of livestock farms. Acknowledgments This research was developed within a wider project, funded by the Umbrian Regional Council, aimed at the evaluation of the environmental sustainability of the pig farming sector in the Trasimeno basin area. The authors wish to thank all the experts who participated in the multicriteria evaluation process, an essential phase in the development of the EASE model. Special thanks go to veterinarians and farmers of the area of Lake Trasimeno for their valuable support and collaboration in data collection. The authors also would like to thank the editors and the three anonymous reviewers for their valuable comments on earlier versions of this article, which helped us greatly improve the quality of the manuscript.
References ANAS (2005) Osservatorio dell’Associazione Nazionale Allevatori Suinicoli. http://www.anas.it. Accessed 23 Jan 2013 (in Italian) APAT (2005) L’inquinamento da nitrati di origine agricola nelle acque interne in Italia. APAT (Italian Agency for Environmental Protection and Technical Services) Report no. 50/2005. http:// www.isprambiente.gov.it/contentfiles/00003700/3766-rapporto2005-50.pdf/at_download/file. Accessed 24 Jan 2013 (in Italian) Bellarby J, Tirado R, Leip A, Weiss F, Lesschen JP, Smith P (2013) Livestock greenhouse gas emissions and mitigation potential in Europe. Glob Chang Biol 19:3–18. doi:10.1111/j.1365-2486. 2012.02786.x Berthiaume P, Bigras-Poulin M, Rousseau AN (2005) Dynamic simulation model of nitrogen fluxes in pig housing and outdoor storage facilities. Biosyst Eng 92:453–467. doi:10.1016/j. biosystemseng.2005.08.008 Berthiaume P, Bigras-Poulin M, Rousseau AN (2007) Sensitivity analysis and application of a dynamic simulation model of nitrogen fluxes in pig housing and outdoor storage facilities. Biosyst Eng 96:455–470. doi:10.1016/j.biosystemseng.2006.12.008 Boggia A, Pennacchi F (eds) (1999) Sviluppo agricolo sostenibile del bacino del Lago Trasimeno. University of Perugia-Dipartimento di Scienze Economiche ed Estimative, ARUSIA-Region of Umbria, Italy (in Italian) Bonazzi G, Fabbri C, Valli L (2003) Allevamenti a basso impatto ambientale. CRPA, Centro Ricerche Produzioni Animali, Regione Emilia Romagna. Informatore Agrario Edizioni, Bologna (in Italian) Boroushaki S, Malczewski J (2010) Participatory GIS: a web-based collaborative GIS and multicriteria decision analysis. Urisa J 22:23–32 Brouwer F (1998) Nitrogen balances at farm level as a tool to monitor effects of agri-environmental policy. Nutr Cycl Agroecosyst 52:303–308. doi:10.1023/A:1009783302364 Burton C, Martinez J (2008) Contrasting the management of livestock manures in Europe with the practice in Asia: What lessons can
1037 be learnt? Outlook Agric 37:195–201. doi:10.5367/0000000087 85915494 Carver SJ (1991) Integrating multi-criteria evaluation with geographical information systems. Int J Geogr Inf Syst 5:321–339. doi:10. 1080/02693799108927858 CRPA (1993) Manuale per la gestione e l’utilizzazione agronomica dei reflui zootecnici. CRPA, Centro Ricerche Produzioni Animali, Regione Emilia Romagna. Tecnograf, Reggio Emilia (in Italian) Degre´ A, Debouche C, Verhe`ve D (2007) Conventional versus alternative pig production assessed by multicriteria decision analysis. Agron Sustain Dev 27:185–195. doi:10.1051/agro: 2007004 Di Fazio S, Modica G, Zoccali P (2011) Evolution trends of land use/ land cover in a Mediterranean forest landscape in Italy. ICCSA 2011, Part I, Lecture notes in computer science, vol 6782, pp 284–299. Springer, Berlin. doi:10.1007/978-3-642-21928-3_20 Duckstein L, Opricovic S (1980) Multiobjective optimization in river basin development. Water Resour Res 16:14–20. doi:10.1029/ WR016i001p00014 Duke JM, Aull-Hyde R (2002) Identifying public preferences for land preservation using the analytic hierarchy process. Ecol Econ 42:131–145. doi:10.1016/S0921-8009(02)00053-8 EEA (2005) Agriculture and environment in EU-15. The IRENA indicator report. EEA (European Environment Agency) Report No. 6/2005, Office for Official Publications of the European Communities, Luxembourg. http://www.eea.europa.eu/publications/ eea_report_2005_6/at_download/file. Accessed 24 Jan 2013 EEA (2007) Europe’s environment. The fourth assessment. EEA (European Environment Agency) State of the Environment Report No. 1/2007, Office for Official Publications of the European Communities, Luxembourg. http://www.eea.europa. eu/publications/state_of_environment_report_2007_1/Belgrade_ EN_all_chapters_incl_cover.pdf. Accessed 24 Jan 2013 EEA (2009) EEA Signals 2009, key environmental issues facing Europe. Office for Official Publications of the European Communities, Luxembourg. http://www.eea.europa.eu/publications/ signals-2009/at_download/file. Accessed 24 Jan 2013 European Commission (2009) The role of European agriculture in climate change mitigation. Commission staff working document, SEC (2009) 1093 final, Brussels. http://ec.europa.eu/agriculture/ climate-change/pdf/sec2009_1093_en.pdf. Accessed 24 Jan 2013 FAO (2009) The state of food and agriculture. Livestock in the balance. FAO, Food and Agriculture Organization of the United Nations, Rome. http://www.globalpolicy.org/images/pdfs/Soc Econ/2010/FAO_food_and_agriculture_report.pdf. Accessed 24 Jan 2013 FAO (2013) Faostat. http://faostat3.fao.org/home/index.html. Accessed 25 Jun 2013 Figueira J, Greco S, Ehrgott M (2005) Multiple criteria decision analysis: state of the art surveys. Springer, New York. doi:10. 1007/b100605 Fusco Girard L, De Toro P (2007) Integrated spatial assessment: a multicriteria approach to sustainable development of cultural and environmental heritage in San Marco deiCavoti. Italy Central Eur J Oper Res 15:281–299. doi:10.1007/s10100-007-0031-1 Fusco Girard L, Torre C (2012) The use of ahp in a multiactor evaluation for urban development programs: a case study. ICCSA 2012, Part II, Lecture notes in computer science, vol 7334. Springer, Berlin, pp 157–167. doi:10.1007/978-3-64231075-1_12 Gerber PJ, Carsjens GJ, Pak-uthai T, Robinson TP (2008) Decision support for spatially targeted livestock policies: diverse examples from Uganda and Thailand. Agric Syst 96:37–51. doi:10. 1016/j.agsy.2007.05.004
123
1038 Gysi C, Schwaninger B (2000) Whole-farm nutrient balance on integrated and organic vegetable farms—a step toward sustainable farming in Switzerland. In Fokkema, Robinson TL (eds) Proceedings of the XXV international horticultural congress, Part 3: Culture techniques with special emphasis on environmental implications, Brussels, Belgium Halberg N, Van der Werf HMG, Basset-Mens C et al (2005) Environmental assessment tools for the evaluation and improvement of European livestock production systems. Livest Prod Sci 96:33–50. doi:10.1016/j.livprodsci.2005.05.013 Hatfield JL, Brumm MC, Melvin SW (1998) Swine manure management. In: Wright RJ, Kemper WD, Millner PD, Power JF, Korcak RF (eds) Agricultural uses of municipal, animal and industrial byproducts. USDA-ARS Conservation. Research Report No. 44 ISTAT (2012) Number of farms and heads. Years 2003, 2005, 2007, 2010. I. Stat, Data warehouse of ISTAT (Italian National Institute of Statistics). http://dati.istat.it/Index.aspx?lang=en. Accessed 24 Jan 2013 Jain DK, Tim US, Jolly R (1995) Spatial decision support system for planning sustainable livestock production. Comput Environ Urban Syst 19:57–75. doi:10.1016/0198-9715(94)00031-X Ju S, DeAngelis DL (2010) Nutrient fluxes at the landscape level and the R* rule. Ecol Model 221:141–146. doi:10.1016/j.ecolmodel. 2009.10.003 Keeney RL, Raiffa H (1993) Decisions with multiple objectives: preferences and value trade-offs, 2nd edn. Cambridge University Press, Cambridge Kesner BT, Meentemeyer V (1989) A regional analysis of total nitrogen in an agricultural landscape. Landsc Ecol 2:151–163. doi:10.1007/BF00126015 Lanyon L, Beegle D (1989) The role of on-farm nutrient balance assessments in an integrated approach to nutrient management. J Soil Water Conserv 44(2):164–168 Lopez-Ridaura S, Van Der Werf H, Paillat JM, Le Bris B (2009) Environmental evaluation of transfer and treatment of excess pig slurry by life cycle assessment. J Environ Manag 90:1296–1304. doi:10.1016/j.jenvman.2008.07.008 Ma J, Scott NR, DeGloria SD, Lembo AJ (2005) Siting analysis of farm-based centralized anaerobic digester systems for distributed generation using GIS. Biomass Bioenergy 28:591–600. doi:10. 1016/j.biombioe.2004.12.003 Malczewski J (2006) GIS-based multicriteria decision analysis: a survey of the literature. Int J Geogr Inf Sci 20:703–726. doi:10. 1080/13658810600661508 Marie M, Srour G, Ziki B (2009) Multi-criteria evaluation of small ruminant farming systems sustainability in Lebanon and Algeria. Options Me´diterr 91:13–20 Martinez J, Dabert P, Barrington S, Burton C (2009) Livestock waste treatment systems for environmental quality, food safety, and sustainability. Bioresour Technol 100:5527–5536. doi:10.1016/j. biortech.2009.02.038 Mendes A, Soares da Silva E, Azevedo Santos J (eds) (2013) Efficiency measures in the agricultural sector. Springer, Dordrecht Modica G, Vizzari M, Pollino M, Fichera CR, Zoccali P, Di Fazio S (2012) Spatio-temporal analysis of the urban–rural gradient structure: an application in a Mediterranean mountainous landscape (Serra San Bruno, Italy). Earth Syst Dyn 3:263–279. doi:10.5194/esd-3-263-2012 Morari F, Lugato E, Borin M (2004) An integrated lpoint source model-GIS system for selecting criteria of best management practices in the Po Valley, North Italy. Agric Ecosyst Environ 102:247–262. doi:10.1016/j.agee.2003.09.025 Munier N (2004) Multicriteria environmental assessment: a practical guide. Kluwer Academic Publisher, Dordrecht
123
Environmental Management (2013) 52:1023–1039 Neri M, Menconi ME, Vizzari M, Mennella V (2010) Propuesta de una nueva metodologı´a para la ubicacio´n de infraestructuras viarias ambientalmente sostenibles. Aplicacio´n en el tramo viario de la pedemontana Fabriano-Muccia. Informes de la Construccion 517:101–112. doi:10.3989/ic.09.043 Nijkamp P, Rietveld P, Voogd H (1990) Multicriteria evaluation in physical planning. Elsevier Science Ltd, Oxford ¨ born I, Edwards A, Witter E, Oenema O, Ivarsson K, Withers PJA, O Nilsson SI, Richert Stinzing A (2003) Element balances as a tool for sustainable nutrient management: a critical appraisal of their merits and limitations within an agronomic and environmental context. Eur J Agron 20:211–225. doi:10.1016/S1161-0301(03)00080-7 Oenema O (2006) Nitrogen budgets and losses in livestock systems. Int Congr Ser 1293:262–271. doi:10.1016/j.ics.2006.02.040 Oenema O, Kros H, De Vries W (2003) Approaches and uncertainties in nutrient budgets: implications for nutrient management and environmental policies. Eur J Agron 20(1–2):3–16. doi:10.1016/ S1161-0301(03)00067-4 ¨ zcan T, C¸elebi N, Esnaf S¸ (2011) Comparative analysis of multiO criteria decision making methodologies and implementation of a warehouse location selection problem. Expert Syst Appl 38:9773–9779. doi:10.1016/j.eswa.2011.02.022 Parris K (1998) Agricultural nutrient balances as agri-environmental indicators: an OECD perspective. Environ Pollut 102:219–225. doi:10.1016/S0269-7491(98)80036-5 Pereira JMC, Duckstein L (1993) A multiple criteria decision-making approach to GIS-based land suitability evaluation. Int J Geogr Inf Syst 7:407–424. doi:10.1080/02693799308901971 Philippe F-X, Cabaraux J-F, Nicks B (2011) Ammonia emissions from pig houses: influencing factors and mitigation techniques. Agric Ecosyst Environ 141:245–260. doi:10.1016/j.agee.2011. 03.012 Piccinini S, Bonazzi G (2005) Nuove strade per smaltire gli effluenti zootecnici. L’Informatore Agrario 61(7):55–60 (in Italian) Regional Council of Umbria, ARPA Umbria (2008) Annuario dei dati ambientali dell’Umbria. http://www.arpa.umbria.it/canale. asp?id=1283. Accessed 22 Jan 2013 Rotz CA (2004) Management to reduce nitrogen losses in animal production. J Anim Sci 82(E. Suppl):E119–E137 Roy B (1968) Classement et choix en pre´sence de points de vue multiples. RAIRO Oper Res 2:57–75 Roy B (1996) Multicriteria methodology for decision aiding. Kluwer Academic Publishers, Dordrecht Saaty TL (1977) A scaling method for priorities in hierarchical structures. J Math Psychol 15:234–281. doi:10.1016/0022-2496 (77)90033-5 Saaty TL (1980) The analytic hierarchy process: planning, priority setting, resource allocation. McGraw-Hill, New York Saaty TL (1994) Fundamentals of decision making and priority theory with the analytic hierarchy process. RWS Publications, Pittsburgh Saaty TL (2004) Decision making—the analytic hierarchy and network processes (AHP/ANP). J Syst Sci Syst Eng 13:1–35. doi:10.1007/s11518-006-0151-5 Saaty TL (2006) Rank from comparisons and from ratings in the analytic hierarchy/network processes. Eur J Oper Res 168: 557–570. doi:10.1016/j.ejor.2004.04.032 Saaty TL, Shang JS (2011) An innovative orders-of-magnitude approach to AHP-based multi-criteria decision making: prioritizing divergent intangible humane acts. Eur J Oper Res 214: 703–715. doi:10.1016/j.ejor.2011.05.019 Saaty TL, Vargas L (2012) Models, methods, concepts & applications of the analytic hierarchy process, 2nd edn. Springer, New York Sangiorgi F, Balsari P, Bonfanti P (1986) Reflui zootecnici. Possibili trattamenti in vista dell’impiego agronomico. Edagricole, Bologna (in Italian)
Environmental Management (2013) 52:1023–1039 Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosales M, de Haan C (2006) Livestock’s long shadow: environmental issues and options. FAO (Food and Agriculture Organization), Rome Steinfeld H, Mooney HA, Schneider F, Neville L (2010) Livestock in a changing landscape. Drivers, consequences, and responses, vol 1. Island Press, Washington DC Sutton Ma, Oenema O, Erisman JW, Leip A, van Grinsven H, Winiwarter W (2011a) Too much of a good thing. Nature 472:159–161. doi:10.1038/472159a Sutton M, Howard C, Erisman J (2011b) The European nitrogen assessment: sources, effects and policy perspectives. Cambridge University Press, Cambridge Teira-Esmatges MR, Flotats X (2003) A method for livestock waste management planning in NE Spain. Waste Manag 23:917–932. doi:10.1016/S0956-053X(03)00072-2 Theobald MR, Dragosits U, Place CJ et al (2005) Modelling nitrogen fluxes at the landscape scale. Water Air Soil Pollut 4:135–142. doi:10.1007/s11267-005-3023-y Tichit M, Puillet L, Sabatier R, Teillard F (2011) Multicriteria performance and sustainability in livestock farming systems: functional diversity matters. Livest Sci 139:161–171. doi:10. 1016/j.livsci.2011.03.006 Torquati B, Vizzari M, Sportolaro C (2011) Participatory GIS for integrating local and expert knowledge in landscape planning. In: Andreopoulou Z, Manos B, Polman N, Viaggi D (eds) Agricultural and environmental informatics, governance and management: emerging research applications. Information
1039 Science Reference, Hershey, pp 378–396. doi:10.4018/978-160960-621-3.ch020 Tress B, Tress G (2003) Scenario visualisation for participatory landscape planning—a study from Denmark. Landsc Urban Plan 64:161–178. doi:10.1016/S0169-2046(02)00219-0 Tzeng G-H, Huang J-J (2011) Multiple attribute decision making: methods and applications. CRC Press, Boca Raton Vargas LG (1990) An overview of the analytic hierarchy process and its applications. Eur J Oper Res 48:2–8. doi:10.1016/03772217(90)90056-H Vincke P (1992) Multicriteria decision-aid. Wiley, New York Vizzari M, Mennella V, Maraziti F (2008) Rischio ambientale nel bacino del lago Trasimeno. Vulnerabilita` del territorio e impatti legati alla gestione dei liquami suinicoli. Faculty of Agriculture, University of Perugia, Perugia (in Italian) Voogd H (1983) Multicriteria evaluation for urban and regional planning. Pion Ltd, London Weiss F, Leip A (2012) Greenhouse gas emissions from the EU livestock sector: a life cycle assessment carried out with the CAPRI model. Agric Ecosyst Environ 149:124–134. doi:10. 1016/j.agee.2011.12.015 Yli-Viikari A, Hietala-Koivu R, Huusela-Veistola E, Hyvo¨nen T, Pera¨la¨ P, Turtola E (2007) Evaluating agri-environmental indicators (AEIs)—use and limitations of international indicators at national level. Ecol Indic 7:150–163. doi:10.1016/j.ecolind. 2005.11.005
123
