THE EU'S CROP DIVERSIFICATION IMPACT ON FLEMISH LANDSCAPE ECOLOGY AND FARM STRUCTURE By MAHY LOUIS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2013
2 2013 Louis Mahy
3 ACKNOWLEDGMENTS I would like to express my very great appreciation to my promotors and tutor, Prof. dr. ir. Jeroen Buysse, Prof. dr. Jeffrey Burkhardt, Prof. dr. John VanSickle and ir. Brnice Dupeux. They provided interesting ideas, as well as the necessary guidance and useful critiques. Also vital for this thesis were Sylvie Danckaert and Joeri Deuninck from the Flemish Ministery of Agriculture, who provided the data. Els Goossens had the patience to read this work and provided useful comments. I also wish to express my gratitude to the jury members. My special thanks are extended to the numerous people behind the IMRD Atlantis project. Without their support, this thesis, the end of two wonderful and challenging years, would not have been possible. I would also like to thank Anne Sophie Holvoet and my family for the everyday support.
4 TABLE OF C ONTENTS P age ACKNOWLEDGMENTS ................................ ................................ ............................. 3 LIST OF TABLES ................................ ................................ ................................ ....... 6 LIST OF FIGURES ................................ ................................ ................................ ..... 7 A BSTRACT ................................ ................................ ................................ ................ 8 1 CROP DIVERSIFICATION IN THE FUTURE CAP ................................ .............. 9 1.1 Problem Analysis ................................ ................................ ......................... 9 1.2 Crop Diversity ................................ ................................ ............................ 13 1.3 Measurement of Crop Diversity ................................ ................................ 17 1.4 Proposed Crop Diversification Measures ................................ ................... 21 1.5 Objective and Research Questions ................................ ........................... 24 1.6 Methodology ................................ ................................ .............................. 25 1.7 Content of the Thesis ................................ ................................ ................ 27 2 METHODOLOGY ................................ ................................ ............................... 29 2.1 Assumptions ................................ ................................ .............................. 29 2.2 Data Pr eparation ................................ ................................ ....................... 33 2.2.1 Selection ................................ ................................ ........................... 33 2.2.2 Re cla ssification ................................ ................................ ................ 35 2.3 The Model ................................ ................................ ................................ .. 37 2.4 Expected Results in Flanders ................................ ................................ .... 39 3 RESULTS ................................ ................................ ................................ ........... 44 3.1 Farm Choices ................................ ................................ ............................ 44 3.1.1 European Commission ................................ ................................ .... 44 3. 1.2 European Parliament ................................ ................................ ...... 45 3.1.3 Council of the European Union ................................ ....................... 47 3.2 Aggregated & Comparative Results ................................ .......................... 48 3.2.1 Impact by Arable Surface ................................ ................................ 48 3.2.2 Crop Mosaic ................................ ................................ ..................... 50 4 CONCLUSION ................................ ................................ ................................ ... 54 4.1 Discussion ................................ ................................ ................................ 54 4.2 Summary ................................ ................................ ................................ ... 57 4.3 Policy Recommendations ................................ ................................ .......... 58 A MEASURE ................................ ................................ ................................ ......... 60 B CROP CLASSIFICATIONS ................................ ................................ ................ 65
5 C POLICY VALUATION ................................ ................................ ......................... 70 C.1 Goals Alternatives Matrix ................................ ................................ ........... 70 C.2 Cost Benefit Analysis ................................ ................................ ................. 72 C.3 Recapitulation and Data ................................ ................................ ............ 75 LIST OF REFERENCES ................................ ................................ ........................... 76 BIOGRAPHIC AL SKETCH ................................ ................................ ....................... 86
6 LIST OF TABLES Table page 1 1 SHDI scores at farm level. ................................ ................................ .................. 19 1 2 Summary of proposals regarding the crop diversification measure. .................... 22 1 3 Overview of the definitions proposed by the European Commission and the Council. ................................ ................................ ................................ ............... 23 2 1 Overview of the scenarios and their respective options and equations. .............. 40 3 1 Changes in crop surfaces compared to the null scenario (2012). ....................... 51 3 2 ............. 52 3 3 and after the simulations. ................................ ................................ .................... 52 3 4 SHDI scores. ................................ ................................ ................................ ....... 52 3 5 Total surface in hectare of leguminous crops, land laying fallow and grassland. 53 B 1 Classification of crops according to the European C ommission and Council proposals. ................................ ................................ ................................ ........... 65 C 1 Example of the goals alternatives matrix. ................................ ........................... 72
7 LIST OF FIGURES Figure page 2 1 Flemish distribution of rotational crop surfaces. ................................ ............................. 42 2 2 (Non )Complying farms, represented by arable surface in Danckaert et al. (2012). ................................ ................................ ................................ ................................ .......... 43 3 1 Farm choices in the European Commission scenario. ................................ ................. 45 3 2 Farm Choices in the European Parliament scenario. ................................ ................... 46 3 3 Farm choices in the Council of the European Union scenario. ................................ 47 3 4 Non complying farms represented by arable surface for all scenarios. .................. 49 3 5 Average number of adopted crops per perturbed farm, represented by arable surface. ................................ ................................ ................................ ................................ ........ 49 3 6 Average proportion of the total eligi ble farm surfaces adapted per perturbed farm, represented by arable surface. ................................ ................................ .................. 50
8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THE EU'S CROP DIVERSIFICATION IMPACT ON FLEMISH LANDSCAPE ECOLOGY AND FARM STRUCTURE By Mahy Louis August 2013 Chair: Robert Jeffrey Burkhardt Major: Food and Resource Economics One of the new, debated political instruments of the upcoming European Common Agricultural Policy reform is the crop diversification measure. European Institutions are currently opposing and dis cussing their respective proposals regarding this greening measure. To comply with the crop diversification measure, arable farmers will have to grow a minimum number of crops on their land, in given proportions. The debate lies in the definitions of crop and the definition of targeted farmers. In this thesis, the central questions are: What are the potential gains in EU farmland biodiversity of the crop diversification measure and to which extend this measure will change farm management practices? Here, we try to predict the impact of the different proposals on farm structures and landscape biodiversity. Changes in ecological values of crop mosaics are modeled, with Flanders as case of crop choice and geographical distance. The outcomes are assessed through the Shannon Diversity Index and several descriptive statistics of farm impact.
9 CHAPTER 1 CROP DIVERSIFICATION IN THE FUTURE CAP The interac tion between agricultural practices and the en vironment is a complex matter It is strongly entangled with contemporary problems as the decline in biodiversity and climate change. Policy makers have started to work on agricultural policies that might address some of the potential problems in the farming environment interaction. In the European Union, this issue has come to be framed in the contex t of the European Common Agricultural Policy of environmentally related c oncerns, among them land us e, water use, soil conservation and biodiversity pre servation (European Commission 2011b). This thesis attempt s to analyze and understand one of the measures of the greening initi ative : crop diversification. This chapter proceeds as follows: In S ection 1, we present an analysis of the problem. Section s 2 and 3 investigate what crop diversity means and how it is measure d S ection 4 describes several proposals for the crop diversification measure. Section 5 present s our research objective and research questions. The methodology is elaborated in S ection 6 a nd t he final section provide s an overview of the content of the thesis. 1.1 Problem A nalysis Agro ecosystems are affected by biodiversity degradation. This causes potential problems for the provision of ecosystem services. Among those are the supporting services, necessary to guarantee provisioning services but also the regulating services, of crucial importance in times of climate change. Agro ecosystems are expected to be influenced by g reater climate variability and changes in precipitation and tempera ture T hese will bring along with them a wide variety of other changes, in cluding changes in the distributio n of species, new disease vectors
10 and potentially harmful invasive species (Lin 2011, FAO 2011). It is therefore crucial to maintain and/or increas e the resilience 1 of agro ecosystems, as an insurance against undesired changes and the resulting welfare losses to farmers and consumers Biodiversity is fundamental to managing the effects of climate change, as i t helps to maintain the stability of ecosy stems and their services (Baumgartner et al. 2009, Perrings 1995, Holling 1973, Tilman 2001, Di Falco et al. 2008). Several institutions have acknowledged the need to improve biodiversity, including the U.S. Congress (Janick et al. 1996), the FAO and the E uropean institutions. These latter are currently working towards a CAP reform which explicitly accounts for the diversity of land use at farm level. Currently, the legislative proposal of the European Commission (EC) is being discussed between the EC, the European Parliament (EP) and the Council of the European Union in view of an expected approval of the revised CAP by the end of 2013. The budget for the CAP was approved in February 2013 with a mandatory greening component going beyon d current cross compliance policy 2 to support agricultural practices deemed beneficial for the climate and the environment. Thirty percent of direct payments to farmers will now be tied to greening These payments will ensure that the farms involved d eliv er environmental and climatic benefits through the retention of soil carbon and grassland habitats associated with permanent pasture 3 the delivery of water and habitat protection by the establishment of ecological focus areas, and the improvement of the r esilience of soil and ecosystems through crop diversification 1 Ecosystem r esilience itself is defined as the capacity to maintain its functions after disturbance (Holling 1973). 2 To receive direct payments, farmers have to respect the rules listed in the Statutory Mandatory Requirements and the Good Agricultural and Environmental Conditions. These are rules related to animal welfare, food safety, plant health and others. 3 Permanent pastures is l and used to grow grasses or other herbaceous forage naturally (self seeded) or through cultivation (sown) and that is not included in the crop rotation of the h olding for five
11 (Europ ean Commission 2011b, Council of the European Union 2013 a, European Parliament 2013 ). In other words, 70% of the single farm payments will continue to be conditional on the existing cross compliance policy while the other 30% will depend on compliance with both the existing cross compliance framework and the additional greening measures. The three greening measures, originally proposed by the EC (European Commission 2011b), are the following. The first crop diversification, is that a arable land 4 will have to be covered by at least three different crops, the first covering a maximum 70% and the third covering a minimum of 5% that arable land These percentages only count for farms where the arable land that is not 100% used for grass production, left fallow or used for crop production under water, nor for land covered with permanent crops or permanent grassland If a arable land is less than three hectare s it does not have to fulfill these obligations to receive the full direct payments Interestingly, t he impact of this measure will be largely influenced by the definition of crop (Matthews 2012). A second measure introduced is the preservation of permanen t grassland. Farmers will have to maintain their surface of permanent grassland declared in 2014. Only 5% flexibility will be allowed, except in the force majeure 2011b, p.40). The last measure is the so EFA s will concern 7% of the arable land, not taking into account permanent grassland. This is described as land left fallow, terraces, landscape features, buffer strips and afforested areas ( more information can be found in European Commission, 2011b). 4 Arable land, as considered b y the European institutions and in this thesis, is distinct from land covered by permanent grassland and crops. Those permanent covers are non rotational and occupy the land for five consecutive years or longer (European Commission 2011b).
12 Although all three measures are related to the health and resilience of agro ecosystems by maintaining or increasing the divers ity of land use in general this thesis focuses on the crop diversification measure. There is nothing new about top field crop known exa mples of this policy approach in history (Tal et al. 2007). The same European institutions that now propose the crop diversification measure have promoted it by including the option for crop rotation measures in the current cross compliance policy Not man y member states chose for this option. Hence, t oday once again, crop diversification is on the This type of policy specifically targets spatio temporal diversity, to have positive impacts on the soil and e cosystem resilience. H omogenization in arabl e crop choices brought on by policies favoring is threatening farmland biodiversity (Bennett et al. 2006, Lin 2011, Benton et al. 2003). Many interlinked factors lie at t he bas is of this homogenization. I ncome stabilizing policies (Di Falco et al. 2005 Bradshaw 2004 ) intensification (Lin 2011), regional and individual specialization (Altieri et al. 2001, Olorode 2007 Bradshaw 2004, Bowler 1985 ) and market liberalization (Fras er 2006 Bradshaw 2004 ) are just some examples They influence the complex cropping decision s of farmers where other factors as yields ( Myers 2009, Bullock 1992) price, labor demand (Myers 2009) and risk management (Bradshaw 2004, Di Falco et al. 2005) play Hence the question remains open whether effort wi ll be sufficient to reverse the homogenization and effectively impact the decline in (farmland) biodiversity (St oddard et al. 2012, Matthews 2012, Westhoek et al. 2012)
13 As previously mentioned, the impact of the crop diversification measure depends to a large extend on the definition of crop. How and which crops are grouped or considered distinct, will determine th e impact on farm management and the landscape (Matthews 2012). Also the inclusion or exclusion of certain farm categories 5 is a point of discussion in the trilogue. The impacts of the proposals from the EC, EP and Council are analyzed from a farm managemen t perspective as well as a landscape perspective. Before elaborating this focus, we provide a short introduction to what is understood by the term crop diversity. 1.2 Crop D iversity In this section we introduce the specific notion of crop diversity as used in this thesis. It is of particular ecological relevance since it is considered the main source of agro biodiversity in intensive production systems (Di Falco et al. 2008, Firbank 2005). Swift et al. (2004) defined global diver sity as the lack of overlap in species, genetic or agroecosystem composition between geographic or temporal s three very important dimensions: Composition: species, genetic or agroecosystem Geographic/Spatial: composition within a geographic boundary at a certain moment in time Temporal: composition over time in a geographic domain. C omposition relates to the building blocks of diversity, and is a crucial aspect of the thesis. The diversity in focus, crop diversity, is al ready a subdivision of general biodiversity H owever regarding the composition of crop diversity there are several levels at which diversity can be assessed, for example: cultivars, speci es, genus, annual/perennial. Different species have distinct impacts o n the broader eco system (McGarigal et al. 201 2, Gardiner et al. 2009), depending on the interactions with 5 The in/exclu sion of farms categorized by size (arable surface) and crops (grassland, leguminous crops, etc.), see section 1.4.
14 oth er components of the ecosystem. In the complex web of an ecosystem, the relations hips among different components can be perceived as functions. That is why much discussion exists as to whether overall species richness is important, or only functional richness. There are arguments in favor of the necessity of a few crucial species performing key roles in the ecosystem functioning (Diaz et al. 2001, Swift et al. 2004). Other s claim that the number of species is more important. An important argument for the latter is that species, a lthough sometimes performing similar roles, through their small differences (Gitay 1996, Brussaard et al. 2007) Redundant or unutilized variety in the present might turn out to be economic value of resilience (Di Falco et al. 2005, Baumgartner et al. 2009) Due to a lack of consensus on the assessment of functional diversity 6 or the methodological difficulties in proposed assessments (Fahrig et al. 2011, Petchey et al. 2006), in this thesis, species richness is used as proxy for proper ecosystem functioning. Much of the former discussion also depends on the spatial scale, the second aspect of crop diversity. Until now many efforts to manage biodiversity, and other environmental issues, are targeted at farm level. However, increased focu s now lies on larger scales, such as the landscape level, since it is at these levels that biodiversity matters more (Swift et al. 2004, Firbank 2005, Fahrig et al. 2011, Thenail et al. 2009 Tscharntke et al. 2005 ). When looking at landscapes as composed of different patches the result is a landscape mosaic (Forman 1995). Consequently, an agricultural landscape is a mosaic where different patches are cropland (Rounse vell et al. 2005), farmers' crop decisions affect landscape mosaics 6 For a complete discussion of this topic see Diaz et al. 2010 and Tilman 2001.
15 significantly. In fact, depending on the dominance of agriculture in a certain landscape or region, the mosaic might be almost exclusively composed of crop patches. Besides, i t is widely a cknowledged that the mosaic structure has an important impact on the ecosystem as a whole (Bennett et al. 2006). The finer the grain of a mosaic the more species are found (Debinski et al. 2001, Duelli 1997). Many of t hese species can provide eco system s ervices to farmers. L arger scale diversified landscapes benefit from a greater ability of pest suppression and reduced pesticide use (Marino et al. 1996, Thies et al. 1999, Benton 2003, Bradshaw 2004, Gardiner et al. 2009). According to the so dive rsity stability hypothesis, more diverse ecosystems benefit from a buffer against climatic shifts or higher variance (Di Falco et al. 2008, Lin 2011 Tilman 200 1 Tscharntke et al. 2005 ). Through niche differentiation 7 the sampling effect 8 and functional diversity they are more stable. Likewise, t he diversity productivity hypothesis claims more productivity fo r diverse ecosystems (Tilman 200 1). Different scales within that mosaic interact and make that the ecologic al value of a crop mosai c is more than the sum of its parts (Benton 2012, Hagedorn et al. 2000, Di Falco et al. 2008, Swift et al. 2004 Gabriel et al. 2010 Tscharntke et al. 2005 ) Examples are within field crop diversity, or the diversity different farm types carry with them 9 They perform vital functions in the ecology of the crop mosaic (Reidsma et al. 2007, Gabriel et al. 2010, Lin 2011). However, to reduce complexity, interactions between crops and other types of land cover will be left out of 7 Niche differentiation: differences in morphology, physiolog y, or behavior of species that can influence their abundances, dynamics and interactions with other species, including the ability of various competing species to coexist (Tilman 2001, p.1) 8 Sampling effect: The hypothesis that diversity might influence an ecosystem process because of the greater chance that a given species trait would be present at higher diversity, and the effect of its presence on ecosystem functioning (Tilman 2001, p.1) 9 As is clarified in Section 1.4, organic farms do not have to comply with crop diversification rule of the EU, the focus of this thesis. This can be considered an argument to leave them out of consideration.
16 consideration. It has been sh own that the mere composition of a crop mosaic has impacts on the ecosystem, with more diversity yielding better results in terms of broader ecosystem diversity ( Bennett et al. 2006, Gardiner et al. 2009, Weibull et al. 2003, Swift et al. 2004, Billeter et al. 2008). Hence for this analysis a macro perspective will be used. An all inclusive perspective at landscape level prevails through this thesis and specific relations between or within the composing elements of the crop mosaic will be left out of con sideration. Besides spatial diversity, the third aspect, temporal diversity also can have beneficial impacts on biodiversity, especially on soil biodiversity (Brussard et al. 2007). This is one of the reasons why crop rotation, growing crops in a recurri ng sequence on the same field, is a widespread practice (Forman 1995). It helps pest and pathogen suppression ( Dick 1992, Krupinsky et al. 2002, IIRR and ACT 2005), and can help to increase productivity (Bullock 1992, Smith et al. 2008). This is because ro t ation reduces the need for many inorganic inputs, certainly if legumes are included in the rotation ( Danckaert et al. 2012 Myers 2009 ). A good crop mixture helps to improve the soil structure (IIRR and ACT 2005). Furthermore, it potentially helps to reta in organic matter (Seremesic et al. 2011) and thus reduces arable lands average carbon emission (Vleeshouwers et al. 2002). This is v ery relevant since an estimated 45% of EU soils have low organic matter which put agro ecosystems productivity in danger (European Commission 2011a). Soil organic matter in turn, is of vital importance for managing water conservation together with plant cover and soil biological activity. It also helps in providing the energy for crucial ecosystem functions (Swift et al. 2 004). The list of possible benefits for soil quality coming from crop rotations is long. The overall result is a better soil resilience (Seybold et al. 1999).
17 In fact, spatial and temporal diversity are strongly related due to the widespread application of rotation. A policy aimed at increasing the spatial diversity of cropland also increases temporal diversity. For example i n the Ge rman measure for crop rotation, promoted through the Euro pean cross compliance policy linked to direct payments there is a re quirement of having three crops on a farm each year (Prinz 2012), nevertheless it remains a crop rotation measure. T he EC summarizes the benefits of spatial crop diversity related to temporal crop diversity through crop rotation, as anic matter (climate change) and structure; reduction of soil erosion and nutrient leaching; nutrients management and input reduction (nutrients and plant protection products); pest and weed control; water quality and quantity; climate change mitigation an d adaptation; improved habitats and landscape This thesis is not about crop rotation as such, but focuses on the first of the three described aspects of diversity It focuses on the changes in composition of the crop landscape caused by the crop diversification measure and the relation to cropland biodiversity The other dimensions are also crucial, as they interact with the composition, but they are assumed to be constant in the sense that the crop diversification mechanism is taken for granted It is analyzed at landscape level for one farm year In this light, we compare t he different proposals of the three different European institutions ( European Commission, Parliame nt, and Council of the European Union ) With the benefits of landscape crop diversity in mind, it is important to take a closer look at how this is measured 1.3 Measurement of Crop D iversity To be effective, the new crop diversification measure s will have to be monitored. This can be done through the land p ar cel identification system part of the
18 Integrated Administration and Control System of the European Union This g eographical information system is already developed by 20 member states (European Commission 2013) and allows the partially automated administration of s Some member states already have an obligatory crop declaration in this system (European Commission 2011a), as it is being used to manage crop rotation comp liance (Westhoek et al. 2012). Crop diversification means that spatial diversity as measured by richness (number of patch types) and evenness (proportion of each patch type) during one crop year is controlled at farm level. It is assumed this will also hav e an effect on temporal diversity of individual plots because of rotation practices. The two aspects of the crop diversification measure, richness and evenness can be measured by the Shannon Diversity Index (SHDI): With m the number of patch types (richness) and P i the proportion of the area covered by patch type i Example s 1 and 3 in Table 1 1 show that the introduction of a new patch type results in a higher SHDI score. For a specific number of patch types the SHDI wil l have its maximum score if all patch types cover an equal share of the area (evenness) (Eiden et al. 2000 ). For example, see the difference between example s 1 and 2 in Table 1 1 The SHDI is a popular measure to assess crop and/or land use diversity (Brad y et al. 2007, Gozdowski et al. 2008, Weibull et al. 2003, Gardiner et al. 2009, Walz 2011 Uuemaa et al. 2009 ). In the case of crop diversification it is also an ideal metric to assess the impact, since both the metric and the implementation scales are ba sed on the relative shares of the different patch types in the relevant area.
19 Table 1 1 SHDI scores at farm level. 1 st Crop 2 nd Crop 3 rd Crop SHDI Example 1 33% 33% 33% 1,09 Example 2 70% 25% 5% 0,74 Example 3 50% 50% 0,69 Other metrics exist as well. The overlap between a well cited principal component analysis (PCA) of landscape metrics (Riitters et al. 1995) and the mosaic concept (Duelli 1997) indicate two important other metrics, the contagion index 10 and the aver age patch perimeter area ratio. These methodological ly and conceptual ly relevant metrics are both correlated to the SHDI. A compositional diverse landscape goes together with more dispersed and interspersed patch types (Riit t ers et al. 1996) More disperse d and interspersed patch types means more patch types per unit area and more possible types of functional relationships between these patch types (Duelli 1997). For landscapes with evenly represented and completely aggregated patch types it can also be sai d that, changes in number of patch types result per definition in changes of the average patch perimeter area ratios. Thus, also if the evenness and aggregation condition s are relaxed, in most cases correlations can be expected. T he average patch perimeter area ratio is relevant because of the ecological importance of edge effects (Duelli 1997, Walz 2011, Bailey et al. 2007, Strand et al. 2007). Besides the correlations with these metrics, there is another reason to choose the SHDI. L andscape mosaic functio nal heterogeneity is a very complex matter. Interactions between different patch types can bring along benefits for one species but not for the other (Gabriel et al. 2010, McGarigal et al. 2012). This is for example why agri environmental schemes in for ex ample the EU and Switzerland are often focused on the habitat production of one species ( Scharntke et al. 2005 Koller 2004, 10 A metric of dispersion and interspersion of th e different patch types in a mosaic (Riitters et al. 1995).
20 Orbicon et al. 2009). A specific measurement excludes certain relations. For example, landscapes with very low average patch perime ter area ratios might boost edge effects, but might inhibit processes depending on core areas 11 (McGarigal et al. 2012). Hence a broadly implemented policy as the crop diversification measure can hardly be evaluated on all types of functional relations and needs an all capturing metric. Bennett et al. (2006) also found that the composition is vital for farmland fauna, especially if the richness of the whole taxonomic assemblage is considered. Hence th e SHDI can serve as a valuable tool for the assessment of the ecolo gical value of the crop mosaic. To be able to use the SHDI, two important aspects of the measurement have to be determ ined. First, crop mosaics as important parts of the landscape will be evaluated at the level of the latter, but what is this landscape level? The range of what is considered a landscape in the literature is wide. However, the bulk of studies work with a nu mber of km ; this fits well with the av erage Flemish community surface of 45,9km ( Agiv 2000, Wiens 1992, Gabriel et al. 2010, Thenail 2009, Weibull et al. 2003). Hence, b y measuring the SHDI at community level we can assume most of the landscape ecologic a l findings to be valid. Second, the validity of the SHDI depends on the classification used to distinguish the different land covers which brings us back to the compositional aspect of crop diversity and its discussion on the importance of species richnes s or functional richness. Due to a missing consensus on an appropriate classification to measure functional diversity (Swift et al. 2004, Diaz et al. 2001, Tilman 2001), the current proxy used is species richness. The SHDI, measured at species and communi ty level s will serve as a proxy to assess the impact of the different proposed crop diversification measures on farmland 11 The interior area of patches.
21 biodiversity The next section presents the different legislative alternatives proposed by the European institutions. 1.4 Proposed Crop D iversification M easure s The EC the Council and the EP have different positions on the future of the CAP. One of the differences concerns the crop diversification measure. The first point of discussion is : which farmers and which parts of their holdings ou ght to be targeted by the crop diversification mea sure? As displayed in T able 1 2 the three European institution s have come up with different ex e m ptions and thresholds. It can be expected that the more farmers become exempted, the smaller the diversifica tion effect will be, although this depends on the nature of the exempted farms. Hence, a first expected result is a trade of between farm management impact and diversity. crop itself. It is a decisive point, relevant for all n on exempted farmers. Here the main alternatives and consequences are presented, followed by the position s and concrete proposals of the different European institutions. The EC provides an overview of the outcomes the definition should benefits in terms of soil organic matter, Commission 2012, p.3). These at a higher taxonomic level ( Stoddard et al. 2012, Westhoek et al. 2012 ). Second, the definition should not threaten the viability of the far ming sector, which could happen if major economic crops are grouped in one crop category. In Flanders for example,
22 Table 1 2 Summary of proposals regarding the crop diversification measure. European Commission European Parli ament Council of the European Union Lower threshold(s) + 3ha : Minimum 3 crops None of 3 < 5% 1 st 10ha to 30ha : Minimum 2 crops + 30ha : Minimum 3 crops 1 st 1 st + 2 nd 10ha to 30ha : Minimum 2 crops + 30ha : Min 3 crops 1 st 1 st + 2 nd Derogation / / Derogation from max. thresholds where main crop is grass or other herbaceous forage Exemption ( s ) Organic farms exempted from greening 100% fallow, grassland, crops under water O rganic farms exempted from greening Organ ic farms exempted from greening + 75% of the eligible area is grassland, cultivated by crops under water, or a combination of those. + 75% of the arable land covered by grass, herbaceous forage, leguminous crops, laying fallow, or a co mbination of tho se + 75% of the eligible agricultural area is covered by Agri Environmental Schemes (Coun cil of the European Union 2005) + 50% of the arable land annually interchanged, each parcel with a different crop than the previous year Source: European Commission 2011a European Parliament 2013 Council of the European Union 2013b
23 grain and forage maize often are planted by the same farmer (Danckaert et al. 2012). Third, the definition should make the system manageable, parcel identification and controls are mentioned. Two controls would be needed for example if a parcel with a green cover sown after harvest would be considered distinct than a parcel without (European Commission 2012). Due to their opposing re quirement s, the goals of environmental benefits and the vi ability of the farm sector are part of a trade off T he different European on this issue translate into some differences in proposed crop classifications (European Commission 2012, Council of the European Union 2 013a, European Parliament 2013 ), which are listed in T able 1 3 below. An impact analysis, as performed in this thesis could bring clarity on the effects of the relation between the different positions, goals and definitions. Table 1 3 Overview of the definitions proposed by the European Commission and the Council. European Commission Coun cil of the European Union Genera Yes Yes Species No Brassicaceae Solanaeceae Cucurbitaceae Gen. Triticum* Winter/Summer No Yes Life cycle No No Other Fallow** Temporary grassland Fallow In this genus and these made at species level, not genera. ** Fallow is considered distinct because of its environmental benefits (European Commission 2012). Source: European Commission 2013 Council of the European Union 2013b Sinc e the proposal is taxonomically more stringent than the s, theoretically it should produce more diversity in a full compliance scenario ( Stoddard et al. 2012, Westhoek et al. 2012 ). Therefore it can be called more environmentally
24 ambitious, bu t it also would have a stronger impact on farm management. For some plant families and one genus distinctions are made at species level. The result is that species such as spelt and wheat (Genus Triticum), rape and colza (Brassicaceae), tomatoes, tobacco and potatoes (Solaneaeceae) are considered distinct. And, more importantly, summer and winter varieties of the same species are also considered distinct by the Council ( Council of the European Union 2012b ). The draft proposal of the European originally proposed another classification, based on a mixture of genera, species and/or varieties (European Parliament 2012) However this definition was not retained, and the fina l EP proposal has delegated authority to the EC to determine the definition of crop (European Parliament 2013) It should be emphasized that these proposals are not yet law (as of June 2013) and are still in a formation process. However, t his thesis hopes to clarify the effects of the different proposals, focusing on the earlier mentioned trade off s To do so, the impact on crop div ersity is investigated, as well as the impact o n farm management. 1.5 Objective and Research Q uestions In this thesis, the central questions are: what is the potential gain in EU farmland biodiversity of the crop diversification measure and to what extent will this measure change farm management practices? Specifically, how do the different proposals of the EC, the European Parliamen t and the Council of the European Union potentially affect farm management decisions regarding crop choices? In short, we try to predict via simulations, the impacts of the different proposals on farm structure and farmland biodiversity. The proposal of a crop diversification measure by the European institutions should be seen in the context of the ongoing CAP reform and the EU biodiversity
25 strategy to 2020. They aim at improving soil and ecosystem resilience. In this approach much depends on what is con sidered a crop and which farmers will have to comply in order to receive fully their direct payments. For each of the examined proposals, the impact on farm management depends on the gap between the present crop configurations at farm level and the requir ements for compliance with the new rules. Therefore, for all the scenarios a full but minimal compliance scenario is computed. The scenarios are full in the sense that all farmers are assumed to comply, yet minimal in the sense that those that have to chan ge their crop configuration to comply only do so to a minimal extents. To model new approach is developed where farmers imitate the crop allocation of the closest peer The closest peer is defined as a farmer already complying with the crop diversification requirements, and who differs the least from the farmer under consideration in terms of crops, their surfaces and geographical distance. Based on the different scenarios, the impacts on fa rmland biodiversity and farm management can be discussed. Estimates concerning complying farmers give a first idea of the impact on the agricultural sector. Issues as differential impacts on different farm sizes are estimated. 12 The SHDI is used to compute a nd compare farmland diversity for each scenario at community level. A higher SHDI score is assumed to be linked with broader farmland biodiversity. The approach taken is explained in the next section. 1.6 Methodology The methodology carrie d out in this stud y involves several steps. First, we assemble data on t he region of Flanders in Belgium which serves as the case study 12 In Danckaert et al. (2012) it was calculated that small farms would be affected stronger than large farms by the crop diversification rule.
26 area The choice for this region is motivated by the existence of an obligatory annual crop declaration for farmers. H ence there is a fu ll coverage of data on farmland use. It is also an area which has some dominant crops which is a problem targeted by the measure under investigation. The raw d ata on Flemish crop land is provided by the Flemish Agenc y for Agriculture and Fisheries This dataset provide s the cover for each agricultural parcel for the year 2012 ( Agentschap voor Landbouw en Visserij 2012 ) After a categorization through G IS software (Esri 2012) by communities (Agiv 2000) and crop categories (A ppendix B ), the data can serve as input for the assessment of the crop diversification measure, independent from the other changes of the proposed CAP reform Next we construct a model that identifies the close s t peer This latter serves as benchmark for the prediction of non complying farmer s reaction To do so the model performs three actions : the first two are the identification of complying farmers, and the iden t ification of the closest peer. The determination of closest peers is based on the Euclidean distance in term s of declared crop surfaces among farms and the geographical distance among their respective communities. To simulate barriers to market entry an additional restriction is built in. O nly peers with maximum two crops more than the non complying farm are cons idered. Also the relative surface of permanent grassland is taken into account, to prevent unrealistic changes with regards to the greening measure on permanent grassland Once the peer farm is determined, the third action of the model is performed which is the projection of the relative crop surface s of that farm on the total surface of the non complying farm. This process is repeated until it can be shown that all farms comply T herefor e the simulation can be called a full but minimal compliance simulati on Furthermore, it was chosen to retain l and that does not need to be
27 diversified (e.g. non rotational covers) to simulate scenarios where farmers adapt to crop configurations exempted from diversification requirements. The use of already complying examp le farms can be perceived as a positive approach because many variables are taken into account if we assume the economic and/or structural optimum of those farms. In other word s farmers will change their crop allocation given the new legis lation while at the same time optimize the economic performance of the farm. An implication of this approach is that we can expect farms to converge to crop configurations that go beyond minimum proportional requirements, and thus beyond minimal compliance. Consequently, this is not an absolute minimum compliance scenario in the strict sense However, it was opted to allow this flexibility because it remains in line with the idea of using viable crop configurations as example s Because of the many uncertaint ies surrounding the delicate process of land allocation at farm level (Thenail et al. 2009) an approach of strictly minimal compliance might simulate non viable allocations. More assumption s can be found in Section 2.1 Finally, the last step is the evalu ation of each scenario using the SHDI at community level, which serves as a proxy for landscapes. Also the number of adapting farms is calculated and the quantity and quality of change. In other words how much land is affected by the change and which cro p s become more or less present? 1.7 Content of the T hesis This thesis is organized as follows. After this introduct ory chapter, the methodology is explained. In this chapter the assumptions are explained followed by a section on the data. It is explained which data is included and how this is included. After this, there is some information on the model used to predict farm behavior and
28 its expected results. Chapter 3 captures the actual results of the model, while chapter 4 contains the conclusion, a short discussion on methodology and the results. The final section of this chapter and the thesis provides some policy re commendations.
29 CHAPTER 2 METHODOLOGY In this chapter, the earlier described approach is more fully elaborated. Starting with the underlying assumptions, some important choices are explained. The positive basis and some normative aspects a re introduced. The following S ection 2.2 explains the steps taken in the data preparation Section 2.3 elaborate s the model and its functioning, while the final S ection 2.4 summarizes some expected results. 2.1 Assumptions Important for all models are the assu mptions made, since they can serve as the building blocks for a model. There are four basic assumptions needed to establish the model used in this thesis. The first assumption, every farmer wants to maximize utility is probably one of the most recurrent assumptions in agricultural economic models. Most often, utility is reduced to profit maximization (Debertin 1993). In this thesis, utility can be left undefined, since no explicit monetary units are used. To maximiz e utility, the farmer has to allocate his resources, in this case, the land to his crops. Here the second assumption comes at play, namely that the observed land allocation is optimal This makes the model positive at its basis (Buysse et al. 2007). The fa rmer considered the relevant variables and made an optimal choice. This assumption is based on the idea of the rational economic man, where optimal choices are made based on objective characteristics and perfect knowledge (Bowbrick 1996). Of course this is impossible to hold to its extremes. However, through this assumption, many variables become endogenous, more variables th an a model explicitly could accommodate A first implication of this second assumption is thus that any imposed deviation of the farme
30 A second is that when a farmer changes his allocation due to the crop diversification rule, he would follow the reasoning of complying farmers, since they have already mad e the optimal choice. Both implications together imply that the diversification. One of the advantages of this mechanism is that by looking at crop combinations, more variables are taken into account than by considering separate crops. Take as an example, two farms with the two main crops wheat and potato e s and one of them with a third cro p, rye Both are probably specialized arable farms, and there is a good chance they have similar soils since they can grow the same crops, the y might have similar machinery and share other similarities In short, the variables associated with the crops are not explicitly considered, but the decision making environment is indirectly simulated. O ne should notic e that a change from the present crop allocation to a new allocation is in fact a violation of the assumption of optimal choice made by farmers. Every farmer has a particular situation, the combination of several factors as soils, risk preferences, capital, location infrastructure brings him to a specific choice in the allocation of his inputs So if through the mo del a new allocation is simulated, certain variables might differ between the particular farm and the farms considered as peer s Therefore, the choices on variables that capture the difference among farms are vital In our model the least differing farm is determined on the following variables: Th e first one is crop surface. Danckaert et al. (2012) showed that the number of crops and/or their proportional allocations on a farm depend on the farm size. More precisely, small farms comply less with the propose d diversification requirements. This argument, together with the fact that policy makers are interested in the impact on
31 different sizes of farms and the structural change of the sector (Buysse et al. 2007) 1 make the choice for absolute crop surfaces reas onable. New crop allocations for no n complying farmers are based on absolute crop surfaces and not relative surfaces, as is the crop diversification rule. Second, as the idea is to simulate full but minimal compliance scenarios, it is also assumed farms on ly want to adopt a minimal number of additional crops and by preference none. This can be linked to entrance and exit costs related to crop productions (European Commission 2011a). Thus the eligible reference farms are limited to those with maximum two c rops more than the original farm These are the two principal variables that determine which farms can serve as reference s However in some cases, there might be several farms eligible. T o differentiate between them, a third variable, the geographical dist ance between is incorporated. A smaller distance goes together with higher chance s of similar soils and transport costs at farm level A third assumption is that changes in land allocations do not affect prices to such an extent th at farmers would consider to shift their production pattern This is Compared to many other models that rely on prices to estimate farm reactions, it would be more complicated to incorporate this type of data since the price is not considered in the objective function. In addition, the simulated region is too small to consider price shift endogenous. The fourth assumption is that all farmers want to comply This is an overgeneralization 2 Claims about the numbe r of farms that would be willing to participate in crop diversification need to be made cautiously. However it can be expected that many 1 Two notes have to be made here. 1) Several proposals increased the lower threshold for the diversification rule, so the smallest farms do not have to comply anymore (Section 1.4). 2) Farms that choose the small farmers scheme, a special measure that regulates the subsidies for small farmers, do not have to comply with the greening measures to receive their full direct payments (European Commission 2011b). 2 It that assumptions need to be realistic.
32 will be rather high. The d irect payments which become partially dependent on the crop d iversification measure, form a considerable amount of the farm income (European Commission 2011d) and for a more demanding greening package than the one actually discussed 3 it has been estimated that in Belgium 117euro would be the total average cost per e ligible hectare. This over estimated cost still lies below 30% of the direct payments per hectare (European Commission 2011c). From this we can infer that it could be compe lling for many farmers to comply with the greening measures, including the crop divers ification measure Costs might differ among the different proposals as it was suggested in Section 1.4. Together with the second assumption this brings us to full, but minimal compliance scenarios. In a pilot run, a large number of farms in the simulatio n raised or lowered their surface of permanent grassland This is unrealistic regarding the environment in which crop diversification will take place. To receive the payments linked to the crop diversification measure farmers also will have to comply with rules regarding permanent grassland (European Commission 2011b, European Parliament 2013 Council of the European Union 2013 a). In other words, lowering a surface of permanent grassland goes against the incentive for crop diversification. On the other hand, increasing the surface of permanent grassland goes together with future, additional limitations of land use, and thus decreasing land prices of the newly classified permanent grassland (Vanoost 2007 in Danckaert et al. 2008). To prevent unrealistic changes in this regard, the choice of reference farms is limited to those with relative surfaces of permanent grassland within a range of 95 to 100% of the same relative surface of the noncomplying farm. 3 Option 3 in the impact assessment of the European Commission, has 10% EFA, 70% green cover, identical crop diversification and permanent grassland requirements (European Commission 2011c).
33 With the first four assumptions and the additional assumption regarding permanent grassland, the basis for the allocation mec hanism in the model is established O ther determinant s for the outcome s are the different crop classifications T he next section explains why this needs some special atte ntion. 2.2 Data Preparation Analysis of the data was divided in to two parts First was the selection of relevant parcels, which is explained in the first subsection. The second part was the categorization of the remaining land covers 2.2.1 Selection The Flem ish Agency for Agriculture and Fisheries provided data on all the agricultural parcels and their cover s in 2012 (Agentschap voor Landbouw en Visserij 2012) This database contains more parcels than needed for this analysis. T he following paragraphs explain the process of selection, based on the relevance of the parcels for crop diversification To determine which farmer holds the land in our reference year, the 21 st of April is used as criterion. This date serves as benchmark for subsidy allocation. Only if a parcel is in use on the 21 st of April is it eligible for direct payments ( Agentschap voor Landbouw en Visserij 201 1) Eligibility for direct payme nts establishes eligibility for greening payments The same date determines the main culture in case of double cropping. All organic farms are excluded from the data, since they are entitled ipso facto to the greening payments (European Parliament 201 3 Cou ncil of the European Union 2013a European Commi ssion 2011 b ). It can be argued that these farms also contribute more to the ecological value of the landscape (Gabriel et al. 2010, Altieri et
34 al. 1996 ) and therefor e are in less need of diversification. They also cover only a very small part of the agricu ltural surface. Crops not eligible for direct payments, such as contains mostly problematic crops. For example, grassland in conservation has to follow specific requirements before being c lassified as such (Agnabio 2013 ). Not all crops can b e used, that is, this part of the landscape mosaic might be in less need of diversification. Third, only a small group of farmers have these crops, so this should not cause any major bias ( Agentschap voor Landbouw en Visserij 2012 ). Other problematic land covers are Agentschap voor Landbouw en Visserij 2011). Although eligible for direct payments, t hese parts of the dataset are best left unconsidered The first two are n ot relevant for a measure related to crop land The last t wo have unknown crops. In case of the small farmer parcels, it can be assumed at least a part of these farmers might choose to participate in the small farmer scheme This makes them ipso facto eligible for greening payments, like organic farms ( European Commission 2011 b ). These four criteria leave us with a crop mosaic consisting of land covered by non rotational and rotational crops, eligi ble for direct payments and relevant to one or more rules of the crop diversification proposals How the different land covers are classified in the model is clarified below.
35 2.2.2 Re classification The different European institutions have proposed different crop classifications, which imply different approaches to attain the effects linked to crop diversificati farmers might or might not face new constraints in their farm management strategies since some crop configurations are not in line with the newly impo sed requirements The model aims to explore reactions It classifies for each scenario the existing cropping pattern and checks if the diversification requirements are met. Unfortunately the crop categories in the original da taset ( Agentschap voor Landbouw en Visserij 2012) and the proposed classifications ( European Commission 2012, Cou ncil of the European Union 2013 ) are not perfectly matched: covered with rotational crops, is distinguished from pe rmanent grassland and permanent crops. These latter have their own regulations in the greening process, so they have a different status in the model. They are exempted from crop diversification requirements because of the link with crop rotation. A conside rable part of the agricultural surface is covered by these permanent covers, for a large part thanks to permanent grassland, which accounts for approximately 23 % of the total Flemish agricultural surface ( Agentschap voor Landbouw en Visserij 2012 ). These non rotational covers were not deleted from the dataset because they remain relevant for the calculation of the SHDI and as optional cropping activities for farms. this category of crops. In t he model they are excluded from the need of crop diversification. The original data has some crop species or genera grouped together F or example are combined and the re
36 are categories or Due to the impossibility of separating them, these crops remain grouped in the simulation, althou gh in reality they should be split into distinct categories Noise caused by this problem is limited since the largest crop that cannot be divided into the correct categories is spelt, which covers only 420 ha (Agentschap voor Landbouw en Visserij 2012). Other crops account for less. On the other hand the original data contains some details not relevant for the other classifications. Distin ctions based on the None of the proposed definitions of crop distinguishes between carrots for fresh markets or industry. For the calculation of the SHDI, the distinction between summer and winter varie ties is also dropped, since species richness is used to determine the ecological value of the crop mosaic ( Section 1.2) In other words, specific cultivars are left out of consideration here It can be assumed the individual crops better represent a farmer decision making environment than the crop classification belonging to the different scenarios. Hence in the model decision making is based on crops as they are commonly perceived (culti vars or species, column 1 in A ppendix B ), and compliance is based on the proposed classifications Besides the assumed correspondence with the relevant aspects of reality, a fine r grained classification also brings the advantage of the possibility of comparing changes in crop ping patterns over scenarios. Additionally there is a more complete representation of allocation strategies, including strategies out of diversification requirements i.e. the switch to permanent crops, different types of grassland, fallow or leguminous crops
37 A complete list of crops present in the or iginal dataset and their respective categories in the different proposals and sc enarios, can be found in A ppendix B How the model deals with these crops is explained in 2.3. 2.3 The Model Based on the assumptions and allocation mechanism described above, this section describes the model. Complying farms remain with the same crop distribution on their holdings. Non complying farms switch to an already existing, complying crop combination, projected on their farm surface. This new crop combination is the one tha t deviates the least from the present crop allocation in terms of crop surfaces and geographical distance. The general requirements for the relationship between representative farms (complying) and constrained farms (in in fringement) are represented by Equ ation 2 1 and Equation 2 3; changes for different scenarios and options within scenarios are discussed below. Consider the following: Equation 2 1 bundles the general requirements on which the choice of closest n farm s, with the possibility to grow c crops. Among those crops, several need to be specified independently as they are related to different rules in the policy packages. Hence p represents permanent grassland, g stands for temporary grassland, h is herbaceous forage, f is the index for fallow and leguminous crops are indexed by l Equation 2 1 identifies the closest peer for farm n referred to as peer Variables are represented by Greek symbols. is a dummy variable wit h value 1 if the conditi ons in Equation 2 2 regarding permanent grassland measure of the greening package are met. The same goes for a dummy variable wi th value 1 if the condition in Equation 2 3 regarding the number of crops is met and where the presence of a crop on a farm is accounted for by a dummy variable. The new crop configuration has by preference not more than 2 crops more than the old
38 configuration The surface allocated to each crop is depicted by the variable Equation 2 1 takes the sum of the absolute differ ences in hectare per crop type between the farms. Finally, represents the geographical distance between the communities of the respective farms, it is introduced to distinguish between farms which have an equal outcome of the former three arguments. Min imize (2 1 ) s.t. ( 2 2) (2 3 ) However, not all farms are eligible as representative farm. Each scenario has some options for compliance. For each of those a subset of equations is added to the general equations where the largest crop on a farm is represented by index 1 the second largest by 2 and the third largest by 3. The variable is introduced to distinguish the arable surface of a farm from the total surface, because the latter also comprises permanent g rassland and permanent crops. It is also necessary to introduce the index to represent farm n after the simulation. More precisely, for 3 (2 4 ) < 0.7 ( 2 5 ) 0.05 (2 6 ) Equation 2 4 represents the requirement related to the total arable surface after the simulation. If the projected arable surface is more than 3 hectare, the eligible peer farms have to comply with the rul es represented in Equation 2 7 and Equation 2 8 The first crop cannot cover more than 70% of the arable surface while the third crop has to cover more than 5% of the arable surface. The other options for
39 compliance are exemptions from diversification requ irements and each contains only one additional equation: to have less than 3 hectare of arable land, 3 ( 2 7) to have the arable surface completely covered by grassland, = ( 2 8) or to have the arable land completely laying fallow = ( 2 9) For each combination of n and peer the right set 4 of equations gives an outcome in E quation 2 1. The combination with the lowest outcome for farm n is the basis for the last step, which is the projection of the new crop configuration on the total surface of the perturbed farm. After this projection formerly not compliant farm become compliant 5 : = ( 2 10) An overview of the options and their respective set of equations can be found in Table 2 1 environmental measures and the interchange of land are not modeled. This is due to uncertainty about future agri environmental measures and ins ufficient data on the interchange of land. Some of the expected results of the different scenarios can be found in the next section, partially inspired on Danckaert et al. (2012). 2.4 Expected R esults in Flanders In this section a short introduction on Flanders and some expected results of the different scenarios are described. Danckaert et al. (2012) already performed an 4 In the EC scenario the possible sets are: 1 6 / 1 3, 7 / 1 3, 8 / 1 3, 9 5 Compliant farms remain the same since they will be the closest peer to themselves.
40 descriptiv e approach similar to theirs is combined with Flemish farm data from 2012 (Agentschap voor Landbouw en Visserij 2012). Table 2 1 Overview of the scenarios and their respective options and equations. Scenario (ha) n / peer requirements Comments European Commission / Farms up to 3ha are exempted > 3 < 0.7 Proportional requirements for first and third crop = / Farms with the arable land completely covered by grassland are exempted = / Farms with the arable land completely laying fallow are exempted European Parliament < 10 / Farms below 10ha are exempted 30 Respective proportional requirements > 30 + Respective proportional requirements Council of the European Union < 10 / Farms below 10ha are exempted 30 or < = Respective proportional requirements, including the derogation where the proportional requirement is relaxed but still two crops need to cover the arable land > 30 + or + < = Respective proportional requirements, including the derogation where the proportional requirement is relaxed but still three crops need to cover the arable land 4/3 ( + + + ) / Exemption for farms where 75% of the arable land is covered by grassland or other herbaceous forage, leguminous crops or laying fallow
41 Flanders is characterized by a high population density : it is strongly urbanized and has a dense road network. Hence, m any stakeholders make claims on the open space (Van Huylenbroeck et al. 2007), which carries a lready a history of small parcels and farms (Vanhaute 2001). To facilitate farm management in this type of fragmented landscape, a series of land consolidation projects have been implemented (Van Huylenbroeck et al. 1996, Van Leirsberge 1991). Although the decline of the farming population resulted in larger farms, the pressure on land prices remains h igh, with an intensive agricultural sector as consequence Although Flanders it accounts for 1,9% of EU production. Approximately 60% of the production value i s made by the livestock sector, while 30% comes fro m horticulture and the remaining 10% comes from specialized arable farm s (Departement Landbouw en Visserij 2006). Since the bulk of the farmers are CAP beneficiaries, greening is relevant to a large part of the agricultural land. More precisely, 24.479 CAP beneficiaries 6 farm 633.412 ha ( Agentschap voor Landbouw en Visserij 2012 ). 145.531 ha are covered by permanent grassland and 14.939 ha are covered by permanent crops. Hence the sion 2011b) The impact on farmland biodiversity strongly depends on the main crops. Figure 2 1 shows that in Flanders over 85% of the non permanent crop surface is covered by five crops: maize, temporary grassland, winter wheat, potato and sugar beets (Agentschap v oor Landbouw en Visserij 2012). Except for wheat, these five major crops are classified equally in the proposed definitions. Therefore, at macro level, only minimal differences can be expected over the proposed definitions. If the other aspects of the prop osals are included, a stronger impact on the overall crop 6 Certified organic farms are excluded.
42 Figure 2 1 Flemish distribution of rotational crop surfaces. ( A gentschap voor Landbouw en Visserij 201 2). If crops are defined by species as th ey are commonly perceived, ( column 1 in A ppendix B ) farms have on average 3,2 crops on their arable land Monocultures constitute 19% of the farms while 24 % have two crops on their arable land Of those monoculture farms, many produce crops belonging to the main crops group. E ven when the list of exemptions is considered, a large number of farmers will have to adopt new crops and de crease their surface of dominant crops. However, due to the nature of the model, it can be expected a substantial part of those newly adopted crops will again belong to those main crops, dampening the former effect on the Also amon g the farmers with three or more crops, 1.489 have a first crop larger than 70% of the arable surface, hence decreasing surfaces of the main crops can be expected. Maize farmers account for 78% of those followed by temporary grassland at 12%. Regarding t emporary grassland, it should be mentioned that in the EC and Council scenarios an increase is also possible, because g rassland farms receive the benefit of exemptions in those proposals and farmers might increase their grassland surface to the necessary level for exemption Maize 41% Temporary grassland 18% Winter Wheat 12% Potato 10% Sugar beets 5% Rest 14%
43 Figure 2 2 (Non )Complying farms, represented by arable surface in Danckaert et al. (2012). Furthermore, Danckaert et al. (2012) included farm size in their analysis. They found a clear relation between the surface of arable land and compliance. Figure 2 2 shows the overrepresentation of not complying farms among smaller farms. Since crop code, a fine grained definition, was used in this analysis these numbers are not representative for the proposals discussed in this thesis. However, they give an insight into this important aspect, and are part of the motivation of the EP and the Council to have special rules for smaller farms. The Council explici tly mentioned this issue and also claims that due to t heir small surface, exemptions for small farms will have a reduced impact on the overall land involved in crop diversification. How much this reduced impact actually is can be derived from the difference between the EC and EP proposal s Th e special regulation for small farmers is the most important point of divergence between those two proposals. In the next chapter we show the results from which conclusions on this issu e and others can be derived.
44 CHAPTER 3 RESULTS This chapter is divided in two sections. First we discuss farm simulated choices with respect to the three proposals. This focuses on the distribution of farms over the options for compliance. Second we clarif y the aggregated results of the farm choices described in the first section. 3.1 Farm C hoices Ordered by resemblance, the simulated crop allocations are presented in this section. First proposal, which has the strongest impact on farm management. 3.1.1 European Commission In this scenario 35% of all farmers need to change their crop allocation to comply. They can do so by ad o pting permanent crops and rotational crops. Hence there is a s mall group of farmers that decreases the surface of the latter by adopting more permanent covers, to become exempted from crop diversification. The group of small farms, with less than 3ha of arable land, increases from 5.098 farms before to 5.245 farms a fter the simulation. Also the exemptions for farms with 100% grassland or fallow on their arable land are used. Among the farms larger than 3ha there are 3 farms where the arable land is completely left fallow after the simulation, instead of only 1 such f arm before the simulation. In the same group the proportion of grassland farms 1 grows from 431 farms to 832, respectively 2,1% to 4,2 %. The other farms above 3ha need to have three crops or more. Before the simulation there are 10.647 complying farms grows to18.769 farms. On the total of 24.839 farms this results in the average adoption of 0,388 crops per farm, this is 1,112 crops per perturbed farm. 1 The farms where temporary grassland covers 100% of the arable land.
45 Figure 3 1 Farm choices in the European Commission scenario. 3.1.2 European Parliament The Euro e (European Parliament 2013a). In the final EP proposal farmers have three options. First, there is the possibilit y to have less than 10ha of arable land and unrestricted crop choices. Before the simulation the number of Flemish farms in this category i s 11.284 ; after the simulation it grows to 11.320 because of the possibility of conversion from arable land to non ro tational covers in the model. In other words 36 farms decreased their surface of arable land to a level below 10ha to become exempted from all further requirements The second option for compliance contains the category of farmers with an amount of arabl e land between 10 and 30ha These Minimum 3 crops 10647 43% 100% grassland or fallow 432 2% Exempted small farmers 5098 20% Non complying farms 35% Complying Farmers 65% Before Minimum 3 crops 18769 76% 100% grassland or fallow 825 3% Exempted small farmers 5245 21% After
46 farmers need to have at least two crops, none of which is allowed to cover more than 80% of the arable farm surface. After the simulation 8.347 farmers belong to this category. The s eventeen percent adapting farms among th ose 8.347 adopted on average 0,92 crops This is lower th an the average increase of 1,13 crops per farm among the adapting farms in the next category, where three crops are required. This last category, where farms have more than 30ha of arable land, counts 5.172 farms after the simulation. 22% of them have a new crop configuration. Over all categories the rate of adapting farms is 11%. Figure 3 2 Farm Choices in the European Parliament scenario. Non complying farms 11% Exempted small farmers a < 10 11284 45% Minimum 2 crops 6869 28% Minimum 3 crops a > 30 3994 16% Complying farms 89% Before Exempted small farmers a < 10 11320 45% Minimum 2 crops 8347 34% Minimum 3 crops a > 30 5172 21% After
47 3.1.3 Council of the European Union The last proposal is distinct from the two former proposals because of a larger number of exemptions and the use of a dif ferent definition of crop. It has the smallest impact on farms as well as on the SHDI. Based on reference year 2012, only 8% of the 24.839 Flemish farms need to change their crop surfaces to comply with the measure. This is for a large part due to the exem ption of farms below 10ha. Also the moderate requirements for farms between 10 and 30ha contribute to this low percentage. E xemptions related to grassland, herbaceous forage, land laying fallow and leguminous crops also decrease the number of farmers not in compliance Figure 3 3 shows the distribut ion of farms over the options. Figure 3 3 Farm choices in the Council of the European Union scenario. Non complying farms 8% Exempted small farmers, a < 10 11284 45% Exempted Derogated 1216 5% Minimum 3 crops a > 30 4006 16% Minimum 2 crops 6447 26% Complying farms 92% Before Exempted small farmers, a < 10 11318 45% Minimum 2 crops 7429 30% Minimum 3 crops a > 30 4665 19% Exempted Derogated 1415 6% After
48 Most of the farms in compliance through an exemption or the derogative rule ( Table 1 2) simultaneously are in line with other options for compliance. Above the first threshold, there are no farms exclusively dependent on the exemption for farms with more than 75% of the eligible are covered by grassland or herbaceous forage, nor exclusively d ependent on the derogative rule for grassland farms. For the derogative rule this changes above the second threshold. Here a small number of farms become dependent on the derogation for compliance, more precisely 288 farms after the simulation. 3.2 Aggregated & Comparative R esults In this section th e resulting impacts on the aggregated crop mosaics and some of the most important differences among the scenarios are highlighted. 3.2.1 Impact by Arable S urface Danckaert et al. (2012) found that the European Commi impact on farms with a larger surface of arable land compared to their smaller counterparts. These findings can be confirmed for the crop data of 2012 ( Agen tschap voor Landbouw en Visserij 2012 ) in combination with the latest proposed crop definition from the EC (European Commission 2012). The other proposals on the contrary, have adapted measures for smaller farms, so their lowest threshold is higher than t measures result in lower, but more evenly distributed proportions of non complying farms over the different ca tegories of arable surface ( Figure 3 4 ).
49 Figure 3 4 Non complying farms represe nted by arable surface for all scenarios. proportion of their farms to become compliant and they adopt a higher number of crops ( Figure 3 5 and Figure 3 6 ). While in the oth er two scenarios, the largest among the adapting farms seem to diverge most from the ir original crop configuration. Figure 3 5 Average number of adopted crops per perturbed farm, represented by arable surface. 0% 10% 20% 30% 40% 50% 60% 70% 80% 3to5 5to10 10to20 20to50 50to100 100+ Arable surface (ha) European Commission European Parliament Council of the EU 0,00 0,20 0,40 0,60 0,80 1,00 1,20 1,40 3to5 5to10 10to20 20to50 50to100 100+ Arable surface (ha) European Commission European Parliament Council of the EU
50 Figure 3 6 Average proportion of the total eligible farm surfaces adapted per perturbed farm, represented by arable surface. The adap ta tions from the EP and the Council for smaller arable farms do have an impact on the number of farms affected and the degree to which they are affected. The ag gregated adapted surfaces can give on the origin of differences in effects on the overall crop mosaic ( next section). Some of the most important statistics: 7.849ha or 22% of the adapted farm surface comes from farms below 10ha If the proportional requirements for farms above 30ha by 7%. The amendment for farms be tween 10 and 30ha from the EP makes the total adapted surface in the res 3.2.2 Crop M osaic Among the expected results was the p rediction that dominant crops would be substituted by other dominant crops. This is the case for all scenarios. Part of the aggregated surface of the most dominant crop after grassland, maize, is substituted by a mix of other cultures, mostly dominant crop s. This effect can be found in all scenarios H owever it is most pronounced proposal followed by the 0% 5% 10% 15% 20% 25% 30% 3to5 5to10 10to20 20to50 50to100 100+ Arable surface (ha) European Commission European Parliament Council of the EU
51 scenario of the EP and Council respectively (Table 3 1 ) The different results for temporary grassland also indicate that the exemptions relat ed to grassland (EC and Council) have an effect on th e aggregated grassland surface. Table 3 1 Changes in crop surfaces compared to the null scenario (2012). Before (ha) European Commission (%) European Parliament (%) Council of the Eur. Union (%) Maize 191270 7 3 3 Permanent grassland 145531 1 0 0 Temporary grassland 84713 +3 1 +2 Winter Wheat 55658 +6 +3 +2 Potato 47809 +8 +4 +2 Sugar beets 22250 +7 +4 +4 Rest 86179 +5 +3 +2 The magnitude of changes is in line with the changes of SHDI sc ores at community level. Table 3 2 shows that the EC scores best. It has the highest mean SHDI score, followed by those of the EP and then the Council. They all are higher than the mean SHDI score before the simulations. The Wilcoxon signed rank test (Tabl e 3 2 ) was performed on the differences between the simulated scenarios and previous situation scenarios. Also shown in Table 3 2 are the effect sizes. Since the data is non normal, is larger while a positive delta indicates the opposite. 1 and 1 are the maximum attainable results. There are considerable overlaps between SHDI scores before and c ontrary to the robust statisti cal values ( Ledesma et al. 20 09 ). This shows that the
52 decrease of maize, the most prominent crop, results in statistically significant increase of eve nness in the Flemish landscapes, but it is less clear whether it has practical significance. Table 3 2 De N Mean Std. Deviation Minimum Maximum EC 304 1,76 0,28 0,49 2,42 EP 304 1,73 0,28 0,52 2,42 Council 304 1,72 0,29 0,52 2,40 Before 304 1,69 0,30 0,52 2,40 Table 3 3 Wilcoxon before and after the simulations. EC Before EP Before Council Before Z 13,856 a 12,272 a 11,784 a Asymp. Sig. (2 tailed) ,000 ,000 ,000 Cliff s delta 0,13 0,07 0,05 a. Based on negative ranks. Another finding depicted in Table 3 4 is that the differences among the scenarios are also statistically significant. Because the species richness remains statistically scenario induce a significant difference. As could be expected from the effect sizes described above, even smaller effects here. Table 3 4 SHDI scores. EC EP EC Council EP Council Z 11,685 b 13,212 b 8,378 b Asymp. Sig. (2 tailed) ,000 ,000 ,000 0,06 0,08 0,02 a. Based on negative ranks.
53 A side note worth mentioning ecological value also depends on the surfaces of specific crops, targeted by the several exemptions and derogations. Grassland, leguminous crops and fallow a re considered ecological ly valuable (European Commission 2012, European Parliament 2010, Lywood et al. 2009 Myers 2009 ) 2 The functions they perform are perceived as vital. The refor specific rules to enhance the respective surfaces can be found in the proposals from the EC and the Council However, there seems to be no in or decrease that is significant in both proportional and absolute terms. A final Table 3 5 shows the total surfaces of these crop types benefiting special regulations. Table 3 5 Total surface in hecta re of leguminous crops, land laying fallow and grassland. Before Commisson Parliament Council Leguminous crops 10.805 12.397 12.132 11.690 Grassland (permanent temporary) 230.245 231.536 228.917 231.964 Fallow 170 290 178 162 2 A considerable part of the added value of leguminous crops is mediated through crop rotation, however information on their presence helps assessing the broader ecological value of the crop landscape.
54 CHAPTER 4 CONCLUSION This final, concl uding chapter contains a discussion on the results and the possible eff ects of methodological choices. A n overall conclusion is delineated, a s well as some policy recommendations. 4.1 Discussion The results illustrate the significant but small effects on SHDI scores of the three proposals. They all have statistically significant impact s on the landscape s Also the differences among the scenarios proved to be significant. The EC results in a higher diversity but it also has the highest number of a ffected farms in all categories of arable farm sizes, and most certainly among the smallest farms. This strong impact on small farmers was the focus of many critiques, recognized by the Council (Council of t he Euro pean Union 2012a). Both the EP and the Council proposed amendments to reduce this differential impact through conditions adapte d to arable surface s In case of the EP, the system with two thresholds reduces the number of non complying farms t o one third of those of the EC. It is uncertain whether this is a general aspect of the proposals, since it should be partially attributed to Flemish low average farm size s (Vanhaute 2001). On top of this, the Council also included derogations and exemptio ns for crop configurations considered beneficial for the environment, containing g rasslands, leguminous crops and others Here the number of non c omplying farms is 8% of all farms, one fourth of the EC where 35% of the farms is non complying Based on the differences in the categories of largest farms some additional conclusions can be drawn. In these categories the difference s between EC and EP
55 are the proportional requirements 1 The EC sets as maximum 70% of the arable land for the first crop a nd minimum 5% for the third while the EP has maximum 75% for the first and 95% for the first two crops. These requirements result in a much higher number of adapting farms in the EC scenario and thus a larger total adapted areal. Although the adapting farms in the 2 which could be interesting from an ecological point of view. Another remark on the largest farms is the trend reversal in average adapted farm surfaces. O nly a small proportion of f arms above 100 hectare are in infringe ment before the simulations, but those perturbed farms change a large part of their farm surface s This might indicate that larger farms might consider large investments instead of using contracting. Those investments should be spread over a large areal. On the other hand, t hese opposing trends might indicate a problem in the model, namely the dependence on closest peers. This is a first important issue with the methodology If the closest peer differs strongly from a given farm, the changes might be over estimated. The model contains a large number of farmers and has much flexibility because of the inclusion of non rotational crops and the possi ble changes in arable surfaces. However, for some farms this is not sufficient to reflect reality. Rare crop conf igurations suffer a higher chance of this type of overestimation since they have less (close) peers. Farms might adopt too many new crops and/or change to proportions far beyond the minim um requirements. For these farms a more normative approach could be followed, where crop configuration projections are no longer based on existing closest peers, but p urely based on the minimum requirements for compliance 1 The exemption for grassland and fallow has negligible effect here. Above 50ha of arable land there are only 10 farms with 100% grassland or fallow on their arable land. Above 100ha, there are none. 2 more complex and cannot be ascr ibed to one specific measure.
56 environment of 2012 w ill be outdated when the crop diversification measure comes into full force This is partly because of the measure itself, but also due to the other novelties of the upcoming CAP reform and the general European economic context Some cropping choices in 20 12 might be come less or more attractive when these new variables are taken into account. This includes coping strategies specific to crop diversification, for example diversification to summer and winter varieties if the Co ssues s, the small farmer scheme and the young farmer scheme all also might come into play (European Commission 2011b, Council of the European Union 2013b ). O ther new factor s are effects mediated through crop diversification and its accompa nying novelties. Take for example auto adjusting market mechanisms. Since the crop diversification measure most likely will be implemented in all EU member states. This might violate the assumption that prices of goods will not change to such an extent the farmer would consider another crop configuration. Over or underestimated crop surfaces are not unthinkable A last issue is related to the unconsidered cropping possibilities T he option of non compliance was not modeled. If farmers consider the greening package too heavy they might choose for non compliance since it is a voluntary measure. The c rop diversification measure seems to induce a higher burden for some farm might be further from reality in those categories. Additionally, two of the options for 3 Hence the reaction of a set of farmers might be different than those simulated. 3 The one for farms with agri environmenatl schemes and for those who interchange their land (Council of the European Union 2013a).
57 Based on these critical element s it can be stated that the estimated changes in SHDI are most probably overestimat ed. The statistically robust but practically less significant effects can be interpreted as directions in which the crop s. 4.2 Summary The decrease in cropland diversity is caused by multiple factors However, awareness rises about the associated environmental risks crop diversification measure is a reaction to this. Many doubts exist on the effects of th is new measure on farmland diversity and its farm level impact ( Stoddard et al. 2012, Westhoek et al. 2012, Matthews 2012) This study has tried to respond to these doubts by elaborating a new approach for the modeling of farm behavior. The result s of this intuitive and strongly positive oriented approach applied on the Flemish case, showed that the implementation of the prop osed diversification mechanism carries the po tential of being a positive alternative to the homogenization of farmland Nevertheless th e effect is to a large extent determined by the specificities of the package of rules The magnitude of change depends on the imposed diversity requirements as well as the criteria of inclusion Due to the different goals associated with the measure, no ne of the proposals of the EC, EP and Council is un equivocally better than the other. T he more stringent and inclusive the policy package the larger the diversity of cropland becomes, but also the greater the burden on farmers. Earlier findings of small less diversified farms were corroborated. Hence the non discriminative EC proposal has a differential impact, smaller farms would suffer the largest burden if the EC proposal was adopted The EP and Council proposals effectively reduce this differential fa rm level burden by targeting through adapted stringency and in/exclusion per arable farm surface
58 claim that the adaptation s for small farms would only slightly affect the overall impact of the crop diversification measure (Council of the European Union 2012a) has shown to be incorrect with respect to Flanders S o m e conclusions can also be drawn regarding the type of landscape diversification. Monocultures are the implicit target of the crop diversification measure and it i s indeed maize, the most dominant crop, which would see the largest reduction in surface. I n this sense, the mechanism is effective. However, the large st part of the freed area was absorbed by other dominant crops Hence, it remains an open question wh ether the diversification effect on farmland is sufficient to reduce the loss in the broader farmland biodiversity. It is advisable to conduct f urther research on this issue on how the crop diversification measure can be refined to correspond with the ser ies of goals it has to meet. 4.3 Policy R ecommendations Since the trade of f between farm management impact and diversity was confirmed, no clear cut policy recommendation can be made. What is cl ear however, is that both a one size fits all policy and its opposite ha ve disadvantages for Flanders. Small farmers are important for the multifunctionality of the European agricultural sector (Guyomard 2004) and due to their plurality they contribute to a higher diversity in the landscape (Rijdsma et al. 2007) To avoid higher burdens in this category and possible perverse effects in a broader multifunctional perspective, a discriminatory diversification approach s amend ments could be advisable On the other hand, diversification requirements for small farmers result in a higher overall diversity. Since there are other ways to achieve soil and ecosystem resilience (Angileri et al. 2011) which possibly yield
59 further research on comple ment ary mechanisms should be conducted A crop diversification mechanism as proposed by the different European institutions has the potential to contribute to more diversity and thus agro ecosystem resilience, but it should not be seen as an isolated sin gular solution for two reasons. First, it only affects a certain type of crop diversity. There are several types of crop diversity with an impact on agro ecosystem resilience. Hence, if resilience is the goal, the discussed crop diversification measure nee ds to be implemented in a package of diversifying measures each targeting the relevant forms of diversity (Lin et al. 2011) 4 Second, t o attain the full effects on the specific type of diversity targeted by the discussed measure, accompanying policies are advisable decision is a complex matter affected by many factors, which all merit attention 5 Furthermore, i t speaks for itself that the important role of the major crops is intrinsic to agro ecosystems that need diversifica tion. In line with Matthews (2012) a suggestion on further research on the role the definition of crop as targeting tool could be made 4 A complete list of crop diversity types can be found in Lin et al. 2011 5 For example, Fras er (2006) showed the relation between crop diversity and the existence of
60 APPENDIX A MEASURE This annex should serve as a broader introduction to the general context in which the crop diversification measure is discussed, part of it is repetition of the first chapter but it gives a completer overview. After this introduction the main points of discussion are described inc luding those beyond the official trilogue The European Common Agricultural Policy is a unified, supranational policy. Contrary to most other fields the agricultural issues are almost exclusively decided at the EU level. In recent years this policy has mo ved towards a more market oriented approach where most of the budget goes to direct payments, assumed to be non distortive and hence welfare maximizing (von Witzke et al. 2010, Moyer et al. 2002) 1 In the single payment scheme farmers receive a payment ba sed on the surface they cultivate. The payment is decoupled from crop type and yield s Through these payments farmers have part of their income guaranteed, independent from market and production risk s To receive these direct payments, farmers have to fu lfill in a series of conditions, called the Good Agricultural and Environmental Conditions and the Statutory Mandatory Requirements, related to environment, animal health and others This is called cross compliance (De Roest et al. 2008). As described in c hapter 1, an environmental extension on this cross compliance is being prepared for the upcoming reform, the greening package. This contains the rules on permanent grassland, crop diversification and EFA 2 Besides this so called greening of the CAP, a ser ies of other changes are in the make. Some are r elevant to greening and thus crop diversification for example 1 A point of discussion, this assumption is contested by many researchers (von Witzke et al. 2010) 2 For a com plete ove rview on the exact proposals : European Commission 2011b, European Parliament 2013, Council of the European Union 2013a
61 the young farmer scheme and the small farmer scheme, where farmers under certain conditions are set free from greening requirements, as are organ ic farms (European Commission 2011b). Also relevant is the proposal of the Council to give member states the possibility to establish a set of environmental certification schemes 3 or they have to be equivalent in practices and attain the environmental goals (Council of the European Union 2013a). This proposed flexibility can be perceived as a p ossibility to introduce crop rotation directly (Matthews 2012). Crop rotation is part of a list of other mechanisms than crop diversification that direct crop rotation i ncentives were rather unsuccessful. Member states had the option to include crop rotation obligations in the cross compliance package. However, only a handful of EU member states implemented them, some on an obligatory basis (e.g. Germany), others optional (e.g. France), some focused on specific crops (e.g. Flanders) others chose for more generic forms (e.g. Scotland). Member states as Austria and Italy have extra rotation payments in the second pillar (Farmer et al. 2004). 4 An important difference lies in tackling rotation direct or indirect. For example Flanders and Austria control the sequence of crops per parcel over the years. Others, like Germany, did this indirectly by having a similar measure to crop diversification. This diversity in policies partly reflects the diversity in ways that can be used to achieve a good overall soil quality. As mentioned earlier crop diversification is only one of many ways (Angileri et al. 2011, IIRR and ACT 2005). 3 For example agri environmental schemes: contractual, long term, voluntary agreements between the farmer and the governments (P olman et al. 2007) 4 For a complete overview of member states applying crop rotation: Farmer et al. (2004).
62 choice for indirect incentives for crop rotation is threefold. One, direct rotation obligations would have a bigger impact (Danckaert et al. 2012, Matthews 2012). With in a crop diversification scheme, a farmer reluctant to practice rotation could maintain the same spatial distribution over several years. Or in case rotation is practice d, the proportional requirements make it possible to keep the same crop on a parcel 2.3 out of 3.3 years 5 Besides this, existing policies will be escribed proposal regarding substituting measures is not accepted more pervasive existing crop rotation rules might disappear, yielding counterproductive effects (Westhoek et al. 2012). Another risk is that if more profound crop rotation measures are moved to the second pillar (Matthews 2012) this might result in inefficiency. Voluntary rules, as the second environmental measures are prone to misallocations of funds (Fraser 2008 ). This is related to the second critique on the mechanism, namel y that farmers managing their soil quality in other ways remain unrewarded (Danckaert et al. 2012, Matthews 2012, Westhoek et al. 2012). The third critique on the mechanism is about the arguments used to implement diversification instead of rotation: admin istrative burden and the WTO. Regarding the former, direct rota tion would rely on the same land parcel identification system ( section 1.3), automated cross annual checks of this data are not impossible; some countries already implemented this (Matthews 201 2). Another important issue is the disconnection from scale. Theoretically a farm surface could have a much higher spatial diversity with a diversity of small farms cultivatin g it (Benton 2012, Rijdsma et al. 2007). Smaller farms also face relatively 5 Calculated with the maximum threshold of 70% for the first crop and the minimum thr eshold of 5% for the third crop from the original EC proposa l.
63 higher cost for the introduction of extra crops. This means a non negligible misallocation that could turn out as a perverse incentive if it would push smaller farms out of busines s. Therefor several demands to increase the lower threshold of 3ha were raised, the Council and E P acknowledged this and proposed to raise the lower threshold (10ha ). Another aspect showing the importance of scale is related to landscape to global homogeni zation of crops (Swift et al. 2004, Firbank 2005). By following diversification requirements, it is not guaranteed that diversity at a landscape level will improve, while it is claimed that diversity at higher levels is more important (Swift et al. 2004). Therefore, the diversity of agro ecological zones should be taken into account (Stoddard et al. 2012, Firbank 2005), and more thorough measures that guarantee sufficient diversity at ecosystem le vel and beyond, should be investigated. mentioned in section Section 2.4 used in the LPIS, the requirements are easily met and diversity will not increase that much (Danckaert et al. 2012). If the requirements are set at a higher taxonomic level, more diversity can be achieved (Stoddard et al. 2012, Westhoek et al. 2012), but at farm level more demanding measures result in high er proposals regarding specific crop types could be seen as an emphasis on functional composition and/or richness (Diaz et al. 2001, Swift et al. 2004). The obligatory introduction of leguminous crop in rotations for example them, is not per se in conflict with WTO rules (Matthews 2012). At the moment the European Council approved the European Multiannual Financial Framework, and the main proposals regarding the CAP are retained. There will b e a flexibility between pillars and the greening component is appro ved.
64 and ought to come to a final decision later this year.
65 APPENDIX B CROP CLASSIFICATIONS This annex contains the table with crop classifications as determ ined by the European Commission (2012) and the Council of the European Union (2013a). Most of the side notes and problems can be found in Section 2.2 of the thesis. Nevertheless, some additional clarifications: column crops are grouped at species level. Permanent crops need no botanical classification in the model since they are only considered as species for the determina tion of the SHDI they are botanically strictly classified classification, thus they are not translated in their Latin names Certain plants/classifications do not have a one to one relation with the original Dutch name and/or scientific name. However none of them causes additional distortion in the model. (Indicated with *). Table B 1 Classification of crops according to the European Commissio n and Council proposals. Species Common European Commission Council of the European Union (Flower) tuber other than begonia (Flower) tuber other than begonia (Flower) tuber other than begonia Alfalfa Medicago Medicago Angelica Angelica Angelica Asparagus Permanent crop Permanent crop Azalea Rhododendron Rhododendron Barley Hordeum Hordeum (Spring) Barley Hordeum Hordeum (Winter) Basil Omicum Omicum Bed and balconyplants Bed and balconyplants Bed and balconyplants Beetroot Beta Beta Begonias Begonia Begonia Blackberry Permanent crop Permanent crop Blackcurrant Permanent crop Permanent crop Blueberries Permanent crop Permanent crop Broccoli Brassica Brassica oleracea Brussels sprouts Brassica Brassica oleracea Buckwheit Fagopyrum Fagopyrum Carrot Daucus Daucus
66 Table B 1 Continued. Species Common European Commission Council of the European Union Cauliflower Brassica Brassica oleracea Celery Apium Apium Celery (var. Dulce) Apium Apium Celery (var. Rapaceum) Apium Apium Chervil Anthriscus Anthriscus Chicory Cichorium Cichorium Chinese cabbage Brassica Brassica Rapa Chive Allium Allium Chrysantemums Chrysanthemum Chrysanthemum Clover Trifolium Trifolium Colza Spring Brassica Brassica Napus (Spring) Colza Winter Brassica Brassica Napus (Winter) Common bean Phaseolus Phaseolus Crisphead lettuce Lactuca Lactuca Cucumber Cucumis Cucumis Cutting plants Cutting plants Cutting plants Early leafy vegetables Early leafy vegetables Early leafy vegetables Eggplant Solanum Solanum melongena Elephantsgrass blessed milk thistle Permanent crop Permanent crop Endive Cichorium Cichorium Fallow Fallow Fallow Fennel Foeniculum Foeniculum Field bean Vicia Vicia Field lettuce Lactuca Lactuca Flax Linum Linum Flowers other than roses Flowers other than roses Flowers other than roses Flowers roses Permanent Crop Permanent Crop Fodder beets Beta Beta Fodder cabbage Brassica Brassica Rapa Fodder carrot Daucus Daucus Fodder rapes Brassica Brassica Napus Forest (short rotation) Permanent crop Permanent crop Forest conifers Permanent crop Permanent crop Forest deciduous ecologic Permanent crop Permanent crop Forest conifers Permanent crop Permanent crop Forest deciduous economic Permanent crop Permanent crop Forest poplar Permanent crop Permanent crop Gooseberries Permanent crop Permanent crop Grapes Permanent crop Permanent crop Grassclover Trifolium Trifolium Green bean Phaseolus Phaseolus Green cellery Apium Apium
67 Table B 1 Continued. Species Common European Commission Council of the European Union Hazelnuts Permanent crop Permanent crop Hemp Cannabis Cannabis Hop Permanent crop Permanent crop Kale Brassica Brassica oleracea Kohlrabi Brassica Brassica oleracea Leek Allium Allium Lettuce* Lactuca Lactuca Lupin Lupinus Lupinus Maize Zea Zea Malting barley Hordeum Hordeum Medicinal and aromatic plants and herbs Permanent crop Permanent crop Millet, sorghum, canary grass or durum wheat Millet, sorghum, canary grass or durum wheat Millet, sorghum, canary grass or durum wheat Mixed grass and leguminous plants (not clover) Mixed grass and leguminous plants (not clover) Mixed grass and leguminous plants (not clover) Mixed leguminous plants Mixed leguminous plants Mixed leguminous plants Non bitter lupin Lupinus Lupinus Noneadible horticultural crops Noneadible horticultural crops Noneadible horticultural crops Oats Avena Avena (Winter) Onion Allium Allium Ornamental trees shrubs Ornamental trees shrubs Ornamental trees shrubs Other annual fruit cultivations Other annual fruit cultivations Other annual fruit cultivations Other berries Permanent crop Permanent crop Other cabbages Brassica Brassica* Other feed crops Other feed crops Other feed crops Other forest Permanent crop Permanent crop Other grains (e.g. meslin) Other grains (e.g. meslin) Other grains (e.g. meslin) Other leguminous green cover Other leguminous green cover Other leguminous green cover Other lettuce Lactuca Lactuca Other non leguminous green cover Other non leguminous green cover Other non leguminous green cover Other oliganeous seeds Other oliganeous seeds Other oliganeous seeds Other ornamental plants Other ornamental plants Other ornamental plants Other perennial fru it cultivations Permanent crop Permanent crop Other spices Permanent crop Permanent crop Other vegetables Other vegetables Other vegetables Paprika Capsicum Capsicum Parsley Petroselinum Petroselinum
68 Table B 1 Continued. Species Common European Commission Council of the European Union Parsnip Pastinaca Pastinaca Peas Pisum Pisum Perennial alfalfa Medicago Medicago Perennial clover Trifolium Trifolium Perennial fruit cultivation (apple) Permanent crop Permanent crop Perennial fruit cultivation (cherry) Permanent crop Permanent crop Perennial fruit cultivation (pear) Permanent crop Permanent crop Perennial fruit cultivation (prune) Permanent crop Permanent crop Perennial grassclover Permanent crop Permanent crop Perennial plants Permanent crop Permanent crop Permanent grassland Permanent crop Permanent crop Phacelia Phacelia Phacelia Potato Solanum Solanum Tuberosum Potato (seedlings) Solanum Solanum Tuberosum Pumpkin Cucurbita Cucurbita* Radish Raphanus Raphanus Rapeseed Winter Brassica Brassica Rapa (Winter) Rapeseed Spring Brassica Brassica Rapa (Spring) Raphanus Raphanus Raphanus Raspberry Permanent crop Permanent crop Red berries Permanent crop Permanent crop Red cabbage Brassica Brassica oleracea Rhubarb Permanent crop Permanent crop Rose Permanent crop Permanent crop Rutabaga Brassica Brassica napobrassica Rye Spring Secale Secale (Spring) Rye Winter Secale Secale (Winter) Rye (Early harvest) Secale Secale (Winter) Salsify Scorzonera Scorzonera Savoy Brassica Brassica oleracea Seedlings non leguminous vegetables Seedlings non leguminous vegetables Seedlings non leguminous vegetables Seedlings Ornamental Permanent crop Permanent crop Shallots Allium Allium Soyabeans Glycine Glycine Spelt Triticum Triticum spelta Spinach Spinacia Spinacia Strawberry Fragaria Fragaria Sugar beets Beta Beta Sunflower seeds Helianthus Helianthus
69 Table B 1 Continued. Species Common European Commission Council of the European Union Tagetes Tagetes Tagetes Temporary grassland Temporary Grassland Temporary Grassland Tobacco Nicotiana Nicotiana Tomato Solanum Solanum lycopersicum Tree breeding Forest Permanent crop Permanent crop Tree breeding Fruit Permanent crop Permanent crop Tree breeding Ornamental Permanent crop Permanent crop Tree breeding Other Permanent crop Permanent crop Triticale x Triticale x Triticale Turf Temporary Grassland Temporary Grassland Turnip Brassica Brassica Rapa Vineyard Permanent crop Permanent crop Walnuts Permanent crop Permanent crop Wheat Spring Triticum Triticum Aestivum (Spring) Wheat Winter Triticum Triticum Aestivum (Winter) White cabbage Brassica Brassica oleracea Willow Permanent crop Permanent crop Winterblooming semi shrub Winterblooming semi shrub Winterblooming semi shrub Winter hardy ornamental plants Winter hardy ornamental plants Winter hardy ornamental plants Yellow mustard Sinapis Sinapis Zucchini Cucurbita Cucurbita Pepo Not translatable, no noise caused.
70 APPENDIX C POLICY VALUATION This impact assessment tried to clarify some of the existing doubts on the decide for the best option, several methodologies exist, among which economic valu ation (Pascual et al. 2007, OECD 2002). The costs and benefits, respectively the farm management impact and biodiversity, can be compared and weighted. The following sections describe two dominant economic evaluation methodologies (Virani et al. 1998). Fir st it is described how the several proposals can be compared on a relative basis, with the goals alternatives matrix, a methodology based on cost effectiveness. The second section describes how the outcome of the first comparison can be further evaluated i n a classic cost benefit analysis. C.1. Goals Alternatives Matrix Among the most prominent valuation techniques is the goals alternatives matrix. In this approach, several policy options can be evaluated on multiple criteria. Each benefit is given a weight, w hich is used to come to the weighted score of benefits. This score is in turn compared to the costs of the respective policy proposal. The result is a score on the cost effectiveness indicator (OECD 2002). How can this be applied on the crop diversificatio n measure and its proposals? The crop diversification measure is associated with a series of benefits (European Commission 2011b). However, in this thesis, only one of those benefits was assessed, the possible impact on biodiversity, which is vital to eco system resilience. Hence only two numbers can be compared for each proposal, the environmental indicator and the cost 1 1 For a full analysis of the policy, the impacts in the other targeted domains should be included as well.
71 Biodiversity indicators try to quantify a concept with a complex definition (Spash et al. 1995). Many types of indicators exist, each wi th their own advantages and disadvantages. The most common is species richness (Gotelli et al. 2001, Diaz et al. 2001). An indicator already described in the thesis, but applied at a different level. The richness, and evenness, of the crop mosaic are assum ed to have positive should be known. Unfortunately, this true effect size has not yet bee n quantified for the SHDI. Costs are conceptually and methodologically easier to define than biodiversity. However, here again, no ready to use data exists to apply on the crop diversification case. Further research is necessary to quantify a monetary valu e of the farm management impact. One should strive for example for an average cost per hectare, or per adopted crop. This can be combined with the respective data obtained in this thesis. The following matrix could clarify what was explained above. Row one depicts the effect sizes of the several proposals on the SHDI, where the goal is to improve ected farmer. As should become clear in the last row, the European Parliament scenario scores best on this fictive cost e ffectiveness indicator. One of the advantages of this approach is that the issue of attributing a monetary value to biodiversity is not necessary. The biodiversity indicator (here crop diversity indicator) accounts for the effect without this type of quantification. Nevertheless, there are two main disadvantages. First, the absolute effect on
72 biodiversity is not considered, the most cost effective scenario does not guarantee a sufficient impact. This issue should be considered explicitly. It is advisable to consider only alternati ves with a sufficient effect, above a certain target (Virani et al. 1998). Second, the most cost effective approach can still be economically uninteresting, more specifically when the costs are larger than the benefits (OECD 2002). A cost benefit analysis performs better on this issue. Table C 6 Example of the goals alternatives matrix. European Commission European Parliament Council of the European Union diversity (SHDI) 0,13 0,07 0,05 Cost Cost effectiveness indicator 1,5 10 5 2,6 10 5 2,5 10 5 C.2. Cost Benefit A nalysis A classic cost benefit analysis has the advantage of comparing both costs and benefits in the same unit. The result is a very comprehensive outcome, one single number indicating the cost benefit ratio. Therefor many researchers and policy makers have tried to apply this on environmental issues, among which biodiversity (OECD 2002, Moons 1999, Ackerman et al. 2001). It can be used to compare several alterna tives, but it also usable for the analysis of a single option. In the case of an environmental policy, expressing the benefits in monetary terms can be of considerable difficulty since environmental goods often lack any market with conventional pricing mec hanisms. This is the case for Flemish biodiversity. It is a public good (Perrings et al. 2003) which means it is non excludable and non rivalrous. All people can enjoy the benefits, in other words enjoy rivalrous means that the benef its for one do not reduce the benefits for another. These characteristics prevent privatization and the use of
73 conventional markets (Polman et al. 2007), hence prices are generally unknown. Second best methodologies have been developed to overcome this qua ntification problem. They can be grouped in three categories (OECD 2002) Revealed preference approaches use existing market values to deduct prices. a) Hedonic pricing b) Travel cost method c) Observed market and related good prices d) Productivity approach e) Cost base d methods, including replacement costs Stated preference approaches use questionnaires to determine the willingness to pay. a) Choice modeling b) Contingent valuation Benefits transfer uses values of already made studies, in other contexts, to determine the value in a specific context. All three categories have their advantages and disadvantages. Due to their nature, revealed preference approaches mainly focus on the instrumental value of biodiversity, for example pest management (Gardiner et al. 2009) or recreational value (Moons 1999). However, many of the instrumental values of biodiversity are unknown, certainly the future instrumental values. Also problematic is that the non use va lues of biodiversity are very significant, the intrinsic values (Nunes et al. 2001, OECD 2002), and these cannot be assessed through revealed preference approaches. In determining the total economic value of biodiversity, including all different values, st ated preference approaches perform much better. Benefits transfer uses studies where revealed or stated preference approaches have been used to determine the value of biodiversity in another context. For example a meta analysis can be conducted. Or a valu e can simply be
74 transposed from one context to the other. Two main issues are important here. First is the availability of the necessary values. Second is the issue of transposing. Transposing can cause noise because of differences in for example socio eco nomic variables which could result in a different value if the transposed study was performed in the new context (OECD 2002). It also requires the same notion of biodiversity. Nunes et al. (2001) clearly show the multitude of notions of biodiversity used i n the literature. This multitude complicates comparison over different studies and areas. With these considerations in mind a methodological choice has to be made. Most researchers use the contingent valuation method for biodiversity (Nunes et al. 2001). T his very straightforward method, where respondents are asked how much biodiversity is worth, includes the intrinsic values and instrumental values (OECD 2002). Important in this approach is that biodiversity is a public good which affects many different st akeholders. Therefore, special attention should go to the representativeness of the study sample. Also important is that knowledge on biodiversity is often precarious, which makes a deliberate valuation complicated (Spash et al. 1995). The necessary inform ation should be provided to improve the validity. The outcome of a contingent valuation study is the estimated willingness to pay per individual. A simple multiplication of this willingness to pay by the size of the population gives the total economic valu e for biodiversity. This has to be weight against the total costs. If the benefits outweigh the costs, welfare is increased and the policy can be considered a good investment. At least, this is the case from a social welfare point of view. Private welfare of certain stakeholders might decrease. This is certainly the case for environmental goods related to agriculture. The bulk of
75 the costs are carried by the farmers and benefits are spread over the population. If the costs outweigh the benefits on the other hand, total welfare is estimated to decrease (OECD 2002). C.3. Recapitulation and Data The two methodologies described above can provide valuable information for the policy decision process. The first, the goals alternatives matrix, can be u sed to compare sev eral policy options. To perform this analysis the following data are needed: analysis on the link between cropland diversity, measured by the SHDI, and agro ecosystem biodiversity, measured by species richne ss, could provide in this. The other option is to perform a n investigation on this link in Flanders. Cropland diversity is well known, data on species richness would need to be assembled. To determine the costs of the policy options, a comparative approach could be followed W here gross margins of diversified, complying farms are compared to less diversified, non complying farm s This way, the multiple economic effects of crop diversification (costs and revenues) are grouped and data to determine the overa ll costs can be deducted (European Commission 2011a) If possible, effects of important variables as farm size and farm sector would have to be controlled for. With these data, the cost effectiveness indicator can be calculated and a comparison among scena rios can be made. To perform the cost benefit analysis, the same data and procedures would be needed. But a measurement of the willingness to pay for biodiversity increase/preservation would need to be added, to quantify the total monetary value of this be nefit The contingent valuation method uses questionnaires to this data. A cost benefit analysis can indicate the economic viability of one single policy scenario, but also compare several policy options. As a final remark, it should be noted that the exp ected effects on biodiversity are not the only impacts For a full economic analysis, also the monetary effects in the other targeted (and non targeted) areas should be included.
76 LIST OF REFERENCES Ackerman, F. and Heinzerling, L. (2001) Pricing the pricesless: cost benefit analysis of environmental protection. Univeristy of Pennsylvania Law Review 150: 1553 1584 Agentscha p voor Landbouw en Visserij (2011) Verzamelaanvraag 2011: Voorwaarden en verplichtingen. Agentschap voor Landbouw en Visserij, Vl aamse Overheid, Brussel. Agentschap voor Landbouw en Viss erij (2012) Algemene Landbouwgebruikspercelen ALV 2012. Agentschap voor Landbouw en Visserij, Vlaamse Overheid, Brussel. Agiv (2000) Voorlopig Referentiebestand Gemeenten. Agentschap voor Geografi sche Informatie Vlaanderen, Vlaamse Overheid, Brussel. AGNABIO (2013). Natuurlijk beheer van grasland. Wat met de steun en randvoorwaarden? Vlaamse Overheid, Brussel. http://lv.vlaanderen.be/nlapps/data/docattachments/natuurlijk beheer grasland.pdf Last accessed 28/04/2013 Altieri, M.A. and Rosset, P.M. (1996). Agroecology and the conversion of large scale conventional systems to sustainable management. Inte rnational Journal of Environmental Studies 50: 165 185. Altieri, Miguel A., and Clara Ines Nicholls (2001). Ecological impacts of modern agriculture in the United States and Latin America. In: Globalization and the Rural Environment. Otto T. Solbrig, Rober t Paarlberg, and Francesco di Castri (eds.). Cambridge: Harvard University Press Angileri, V., Loudjani, P. and Serafini, F. (2011) GAEC implementation in the European Union: situation and perspectives. Italian Journal of Agronomy 6(s1): e 2. Bailey, D., Billeter, R., Aviron, S., Schweiger, O. and Herzog, F. (2007) The influence of thematic resolution on metric selection for biodiversity monitoring in agricultural landscapes. Landscape Ecology 22(3): 461 473 Baumgrtner, S. and Strunz, S. (20 09) The economic insurance value of ecosystem resilience. Department of Sustainability Sciences, Leuphana University of Lneburg, Germany. Bennett, A.F., Radford, J.Q. and Haslem, A. (2006) Properties of land mosaics: Implications for nature conservation in agricultural environments. Biological Conservation 133: 250 264 Benton, T.G., Vickery, J.A. and Wilson, J.D., (2003). Farmland biodiversity: is habitat heterogeneity the key? Trends in Ecology and Evolution 18: 182 188. Benton. T.G. (2012) Managing a gricultural landscapes for production of multiple services: the policy challenge. International Agricultural Policy 1 : 7 17.
77 Billeter, R., Liira, J., Bailey, D., Bugter, R., Arens, P., Augenstein, I., Aviron, S. et al. (2008). Indicators for biodiversity in agricultural landscapes: a pan European study. Journal of Applied Ecology 45 : 141 150 Bowbrick, P. (1996) Quality theories in agricultural economics. European Association of Agricultural Economists: Meeting on Product Quality. Wageningen University, Netherlands. Bowler, I., (1985) Agriculture under the CAP: A Geography. Manchester University Press, Mancheste r. Bradshaw, B. (2004) ? Questioning crop diversification as a response to agricultural deregulation in Saskatchewan, Canada. Journal of Rural Studies 20: 35 48 Brady, M., Kellermann, K., Sahrbacher, C. and Jelinek, L. (2009) Im pacts of decoupled agricultural support on farm structure, biodiversity and landscape mosaic: Some EU Results Journal of Agricultural Economics 60 (3): 563 585 Brussaard, L., de Ruiter, P.C. and Brown, G.G. (2007) Soil biodiversity for agricultural sust ainability. Agriculture, Ecosystems and Environment 121: 233 244 Bullock, D.G. (1992) Crop Rotation. Critical Reviews in Plant Sciences 11(4): 309 326 Buysse, J., Van der Straeten, B., Nolte, S., Claeys, D. and Lauwers, L. (2011) Optimisation of implem entation policies of environmental quota trade. European Review of Agricultural Economics, 1 19 Council of the European Union (2005) Council Regulation (EC) No 1698/2005 of 20 September 2005 on support for rural development by the European Agricultural F und for Rural Development (EAFRD). Brussels, 20/09/2005. Council of the European Union (2012a) Proposal for a Regulation of the European Parliament and of the Council establishing rules for direct payments to farmers under support schemes within the framework of the common agricultural policy (CAP Reform) Commission services Concept Paper on Green ing 2011/0280 (COD) Brussels, 11/05/ 2012. Council of the European Union (2012b) Draft Regulation of the European Parliament and the Council establishing rules for direct payments to farmers under support schemes within the framework of the common agricult ural policy. Working Document 10890/12. Brussels, 8/6/2012 Council of the European Union (2013a) Proposal for a Regulation of the European Parliament and of the Council establishing rules for direct payments to farmers under support schemes within the framework of the common agricultural policy (CAP Reform) Presidency consolidated draft Regulation 7183/13. Brussels 12/3/2013
78 Council of the European Union (2013b) Proposal for a Regulation of the European Parliament and of the Council establishing r ules for direct payments to farmers under support schemes within the framework of the common agricultural policy (CAP Reform) Preparation of the first informal trilogue 7703/13. Brussels 22/3/2013 Danckaert S., Carels K. and Van Gijseghem D. (2008) J uridisch wetenschappelijke toestand van blijvend grasland in Vlaanderen in het kader van de randvoorwaardenregeling, Departement Landbouw en Visserij, afdeling Monitoring en Studie, Brussel Danckaert S., Deuninck J., Van Gijseghem D. and Oeyen A. (2012) Midterm evaluatie van het Vlaams Programma voor Plattelandsontwikkeling 2000 2006. Ministerie van de Vlaamse Gemeenschap, Administratie Land en Tuinbouw, Vlaamse Onderzoekseenheid Land en Tuinbouweconomie, Brussel. De Roest, K., Jongeneel, R., Dillen, K and Winsten, J. (2008) Cross compliance and competitiveness of the European Beef and Pig Sector. Paper prepared for the 109th EAAE Seminar. Wageningen University. Debertin, D. L. (1993) Agricultural Production Economics New York: Macmillan, Inc. 1986. A graduate level agricultural production economics text. Reviewed in American Journal of Agricultural Economics May, 1 987. Debinski, D.M., Ray, C. and Saveraid, E.H. (2001). Species diversity and the scale of the landscape mosaic: do scales of movement and patch size affect diversity? Biological Conservation 98(2): 179 190. Di Falco, S., and Chavas, J.P. (2008). Rainfall shocks, resilience, and the effec ts of crop biodiversity on agroecosystem productivity. Land Economics 84(1): 83 96. Di Falco, S. and Perrings, C. (2005) Crop biodiversity, risk management and the implications of agricultural assistance. Ecological Economics 55: 459 466 Daz, S. and Cabido, M. (2001) Vive la diffrence: plant functional diversity matters to ecosystem processes Trends in Ecology and Evolution 16: 646 655. Dick, R.P. (1992) A review: long term effects of agricultural systems on soil biochemical and microbial paramete rs. Agriculture, Ecosystems & Environment 40(1 4 ) : 25 36 Duelli, P. (1997) Biodiversity evaluation in agricultural landscapes: an approach at two different scales Agriculture Ecosystems and Environment 62: 81 91. Eiden, G., Kayadjanian, M and Vidal, C. (2000) Capturing landscape structures: Tools. In: From Land Cover to Landscape Diversity in the European Union. DG AGRI, EUROSTAT and the Joint Research Centre (Ispra) and th e European Environmental Agency, Brussels.
79 Esri (2012) ArcGIS Desktop: Release 10.1. Redlands, CA: Environmental Systems Research Institute. European Commission (2004) Commission Regulation (EC) No 796/2004 of april 2004 laying down detailed rules for implementation of cross compliance, modulation and the integrated administration and control system provided for in of Council Regulation (EC) No 1782/2003 establishing common rules for direct support schemes under the common agricultural policy and establishing certain support schemes for farmers, pbl, 141, 30/4/2004. European Commission (2011a). Impact assessment Common Agricultural Policy towards 2020 Commission staff working paper. SEC(2011) 1153, pp. 85. European Commission, Brussels, 12/10/2011 European Commission (2011b). Proposal for a Regulation of the European Parliament and of the Council establishing rules for direct payment schemes for farmers under the common agricultural policy, COM(2011) 625/3, Brussels. European Commission (2011c). Impact Assessment Common Agricultural Policy towards 2020. Annex 2D: Greening results of partial analysis on impact on farm income using FADN. European Commission, Brussels. European Commission (2011d). Impact Assessment Common Agricultural Policy towards 2020. Annex 3A: Greening results of partial analysis on impa ct on farm income using FADN. European Commission, Brussels. European Commission (2012) CAP Reform. Fiche No 15. Direct Payments File: 2011/028 0 (COD). Brussels, 19/4/201 2. European Commission. (2013) Land Parcel Identification System. Joint Research Centre, Institute for Environment and Sustainability. http://marswiki.jrc.ec.europa.eu/wikicap/index. php/LPIS Last accessed 28/04/2013 European Council (2013) 7/8 February 2013 Conclusions (Multiannual Financial Framework) EUCO 37/13. Brussels, 8/2/2013 European Parliament (2010) Draft Report on the EU protein deficit: what solution for a long stan ding problem? (2010/2011(INI)). Committee on Agriculture and Rural Development. Brussels, 29/10/ 2010 European Parliament (2012) Draft report on the proposal for a regulation of the European Parliament and of the Council establishing rules for direct payme nts to farmers under support schemes within the framework of the common agricultural policy (COM(2011)0625final/2 C7 0336/2011 2011/0280(COD)) Committee on Agriculture and Rural Development Rapporteur: Luis Manuel Capoulas Santos. Brussels, 30/05/2012
80 European Parliament (2013) Proposal for a European Parliament decision on the opening of, and mandate for, interinstitutional negotiations on the proposal for a regulation of the European Parliament and of the Council establishing rules for direct paymen ts to farmers under support schemes within the framework of the common agricultural policy. Plenary sitting 4/2/2013. Fahrig, L., Baudry, J., Brotons, L., Burel, F. G., Crist, T. O., Fuller, R. J., Sirami, C., Siriwardena, G. M. and Martin, J. L. (2011), Functional landscape heterogeneity and animal biodiversity in agricultural landscapes Ecology Letters 14: 101 112. FAO (2011) Climate Change and Genetic Resources for Food and Agriculture. Commission on Genetic Resources for Food and Agricul ture: Thirteenth Regu lar Session. Rome, 18 22/7/2011. Farmer, M. and Swales, V. (2004) The development and implementation of cross compliance in the EU 15: an analysis. A Report for the RSPB. Institute for European Environmental Policy London Firbank, L.G. (2005) Striking a new balance between agricultural production and biodiversity. Annals of Applied Biology 146: 163 175 Forman, R.T.T. (1995) Land mosaics: The ecology of landscapes and regions. Cambrige University Press, UK, Cambridge Fraser, E.D. G. (2006) Crop diversification and trade liberalization: Linking global trade and local management through a regional case study. Agriculture and Human Values 23:271 281 Fraser, R. (2008) Land heterogeneity, agricultural income forgone and environmental benefit: An assessment of incentive compatibility problems in environmental stewardship schemes. AES, 82nd Annual Conference 2008 RAC, Cirencester 31st March to 2nd April 2008 Gabriel, D., Sait, S.M., Hodgson, J.A., Schmutz U., Kunin, W.E. and Benton, T. G. (2010) Scale matters: the impact of organic farming on biodiversity at different spatial scales. Ecology Letters 13 (7): 858 869 Gardiner, M.M., Landis, D.A., Gratton, C., DiFonzo, C.D., O Neal, M., Chacon, J.M. et al. (2009). Landscape diversity enhances biological control of an introduced crop pest in the north central USA. Ecology Applications 19(1): 143 54 Gitay, H., Wilson, J.B. and Lee, W.G. (1996) Species redundancy: a redundant concept? Journal of Ecology 84: 121 124 Gotelli, N.J. and Co lwell, R.K. (2001) Quantifying biodiversity: procedures and pitfalls in the measurement and comparison of species richness. Ecology Letters 4(4): 379 391 Gozdowski, D., Roszkowska gmina level and its causes in the Podlasie district of Poland. Communications in Biometry and Crop Science 3 (2): 72 79
81 Guyomard, H., Le Moul, C. and Gohin, A. (2004) Impact of alternative agricultural income support schemes on multiple policy goals. European Review of Agri cultural Economics 31(2) : 125 148 Hagedorn, K., Arzt, K. and Peters, U. (2002). Institutional arrangements for environmental co operatives: a conceptual framework. In Hagedorn, K. (ed.), Environmental Co operation and Institutional Change: Theories and Pol icies for European Agriculture Cheltenham, UK: Edward Elgar, 3 25. Holling, C.S. (1973). Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4: 1 23. IIRR and ACT. (2005). Conservation agriculture: A manual for farmer s and extension workers in Africa International Institute of Rural Reconstruction, Nairobi; African Conservation Tillage Network, Harare. Janick, J., Blas M.G., Johnson, D.L., Jolliff, G.D., and Myers R.L. (1996). Diversifying U.S. crop production. In: J Janick (ed.), Progress in new crops ASHS Press, Alexandria, VA. Koller, N (2004) Revue suisse d'agriculture. 36(2): 65 69 Krupinsky, J.M., Tanaka, D.L., Merrill, S.D., Liebig, M.A., Hendrickson J.R., Anderson, R.L., Hanson, J.D. and Ries, R.E. (2002). Crop sequences influence crop seed production and plant disease. p. 13 19. In Proc. of the 22nd Annual Manitoba North Dakota Zero Till Conference, Minot, ND. 29 30 Jan. 2002. Ledesma, R., Macbet h, G. & Cortada de Kohan, N. (2009). Computing effect size measures with ViSta the visual statistics system. Tutorials in Quantitative Methods for Psychology 5(1): 25 34 Lin, B.B. (2011). Resilience in Agriculture through Crop Diversification: Adaptive Management for Environmental Change. Bioscience 61: 183 193. Longman dictionary of contemporary English (2001). Harlow, England: Longman. Lywood, W., Pinkney, J., and Cockerill, S. (2009) Impact of protein concentrate coproducts on net land requirement for European biofuel production. GCB Bioenergy 1: 346 359. Marino ,P.C. and Landis, D.A. (1996). Effect of landscape structure on parasitoid diversity and parasitism in agroecosystems. Ecological Applications 6: 276 284. Matthews, A. (2012) Environmental pub lic goods in the new CAP: Impact of greening proposals and possible alternatives. Directorate General for Internal Policies. Policy Department B: Structural and Cohesion Policies, Agriculture and Rural Development. Brussels, European Union
82 McGarigal, K., Cushman, S.A. and Ene, E. (2012) FRAGSTATS v4: Spatial Pattern Analysis Program for Categorical and Continuous Maps. Computer software program produced by the authors at the University of Massachusetts, Amherst. Moons, E. ( 1999 ) Estimation of the recreat ion values of a forest, M.Sc. dissertation, Faculty of economics and applied economics, Catholic University of Leuven M oyer, W. and Josling, T. (2002). Agricultural Policy Reform: Politics and Process in the EU and US in the 1990s Global Environmental Governance Series, Aldershot; Burlington, VT. Myers, R (2009) Diversifying agriculture with alternative crops. In: Hudson, R.J. (ed) Management of Agricultural, Forestry and Fisheries Enterprises. Vol II. EOLSS e book and print by U NESCO Encyclopedia of Life Support Systems. Nunes, P. and van den Bergh, J. (2001) Economic valuation of biodiversity: sense or nonsense? Ecological Economics 39(2): 203 222 OECD (2002) Handbook of biodiversity valuation: a guide for policy makers. OECD, Paris Olorode, O. (2007) Biodiversity, globalization and poverty. African Journal of Traditional, Complementary Medicine 4(4): 532 540 Orbicon, Atecma, cosphre and Ecosystems (2009) Wildlife and sustainable farming initiative: Good practice examples of how to support species conservation through rural development programmes. Prepared for the European Commission November 2009 Directorate General for Environment. Pascual, U. and Perrings, C. (2007) Developing incentives and economic mechanisms for in situ biodiversity conservation in agricultural landscapes. Agriculture, Ecosystems & Environment 121(3): 256 26 8 Perrings, C. (1995). Biodiversity conservation as insurance. In T. Swanson (ed.), Economics and Ecology of Biodiversity Decline. Cambridge University Press. Perrings, C. and Gadgill, M. (2003) Conserving biodiversity: Reconciling local and gloval publ ic benefits. In: Kaul, I., Conceicao, P., Le Goulven, K. and Mendoza, R. (ed). Providing Global Public Goods. Managing Globalization, United Nations Development Programme. Oxford University Press, New York. Perrings, C., Jackson, L., Bawa, K., Brussaard, L ., Brush, S., Gavin, T., Papa, R., Pascual, U., and De Ruiter, P. (2006) Biodiversity in agricultural landscapes: Saving natural capital without losing interest. Conservation Biology 20 (2): 263 264. Petchey, O.L. and Gaston, K.J. (2006) Functional divers ity: back to basics and looking forward. Ecology Letters 9: 741 758
83 Polman, N.B.P. and L.H.G. Slangen (2007) The design of Agri environmental schemes in EU: Lessons for the future. ITAES report. Integrated Tools to design and implement Agro Environmental Schemes (ITAES). Prinz. C.: GAEC: Progress in implementation in Germany Federal Ministry of Consumer Protection, Food and Agriculture EC Budget, CAP Financing, Control Matters. http://www.ireas.cz/download/projekty/www_cro/presentation/blok3/prinz_ger many.pdf Last accessed 9/12/2012 Reidsma, P., and Ewert F. (2008). Regional farm diversity can reduce vulnerability of food production to climate change. Ecology and Society 13 (1): 38. S.P., Jones, K.B. and Jackson, B.L. (1995 ) A factor analysis of landscape pattern and structure metrics Landscape Ecology 10(1) . A note on contagion indices for landscape analysis. Landscape Ecology 11(4): 197 202 Rounsevell, M.D.A., Ewert, F., Reginstera, I., Leemans, R. and Carter, T.R. (2005) Future scenarios of European agricultural land use. Projecting changes in cropland and grassland. Agriculture, Ecosystems and Environment 107: 117 135 Seremesic, S., Milosev, D, Djalo vic, I., Zeremski. T. and Ninkov. J. (2011) Management of soil organic carbon in maintaining soil productivity and yield stability of winter wheat. Plant Soil Environment 57 (5): 216 221 Seybold C.A., Herrick, J.E. and Brejda J.J. ( 1999) Soil resilience: A fundamental component of soil quality Soil Science 164: 224 234 Smith, R.G., Gross, K.L. and Robertson, G.P. (2008) Effects of crop diversity on agroecosystm function: Crop Yield Response. Ecosystems 11: 355 366 Spash, C.L. and Hanley, N. (1995) Methodological and ideological options: Preferences, information and biodiversity preservation. Ecological Economics 12(3): 191 208 Stoddard, F. and Murphy Bokern, D. (2012) Reform of the Common Agricultural Policy to improve the envir onment. A Statement from the 12th Congress of the European Society for Agronomy. ESA European Society for Agronomy, France: Montpellier Strand, H., Hft, R., Strittholt, J., Miles, L., Horning, N., Fosnight, E. and Turner, W., eds. (2007). Sourcebook on r emote sensing and biodiversity indicators: Prepared by the NASA NGO Biodiversity Working Group and UNEP WCMC to support implementation of the Convention on Biological Diversity CBD Technical Series, 32, Montreal (Secretariat of the Convention on Biologica l Diversity).
84 Swift, M.J., Izac, A. M.N. and van Noordwijk, M. (2004) Biodiversity and ecosystem services in agricultural landscapes are we asking the right questions? Agriculture, Ecosystems and Environment 104 (2004) 113 134 Tal, A. and Cohen, J.A. ( 2007) Bringing top down to bottom up: A new role for environmental legislation in combating desertification. Harvard Environmental Law Review 31. Thenail, C., Joannon, A., Capitaine, M., Souchere, V., Mignolet, C., Schermann, N., Di Pietro, F., Pons, Y., Gaucherel, C., Viaud, V., Baudry, J., et al. (2009). The contribution of crop rotation organization in farms to crop mosaic patterning at local landscape scales. Agriculture, Ecosystem and Environment 131: 207 219. Thies C. and Tscharntke T. (1999). Lands cape structure and biological control in agroecosystems. Science 285: 893 895. Tilman, D. (2001). Functional diversity. In: Levin, S. A. (Ed.) Encyclopedia of Biodiversity Volume 3. Academic Press. Tscharntke, T., Klein, A.M., Kruess, A., Ingolf, S. and T hies, C. (2005) Landscape perspectives on agricultural intensification and biodiversity ecosystem service management. Ecology Letters (8): 857 874. Uuemaa, E., Antrop, M., Roosaare, J., Marja, R. and Mander, (2009) Landscape metrics and indices: An overview of their use in landscape research. Living Reviews in Landscape Research 3(1). Van Huylenbroeck, G., Coelhot, J.C. and Pinto, P.A. (1996) Evaluation of land consolidation projects (LCPs): A multidisciplinary approach. Journal of Rural Studies 12(3): 297 310 Van Huylenbroeck, G., Vandermeulen, V., Mettepenningen, E. and Verspecht, A. (2007) Multifunctionality of agriculture: A review of definitions, evidence and instruments. Living Reviews in Landscape Research 1(3) Van Leirsberge, H. (1991) Berekening van de landbouweconomische baten van de ruilverkaveling van bouwland. Masters Dissertation, Rijksuniversiteit Gent, Koninklijke Universiteit Leuven. VanHaute, E. (2001) Rich agriculture and poor farmers: Land, landlords and farmers in Flander s in the eighteenth and nineteenth centuries. Rural History 12 (1): 19 40. Vanoost, K. (2007) Permanent pastures ratio: experiences in Flanders, Agentschap management, control, advisory sup 2007, Ispra.
85 Virani, S. and Graham, S. (1998). Economic evaluation of environmental policies and legislation (Final Report prepared for European Commission Directorate General III (Industrial Affairs) by Risk & Po licy Analysts Limited, Project: Ref/Title J236/CBA/CEA). Brussels: European Commission. Vleeshouwers L. M. and Verhagen, A. (2002). Carbon emission and sequestration by agricultural land use: a model study for Europe. Global Change Biology 8: 519 530. von Witzke, H., Noleppa, S. and Schwarz, G (2010) Decoupled payments to EU farmers, production, and trade: An economic analysis for Germany. Working Paper, Humboldt University of Berlin Walz, U. (2011) Landscape Structure, Landscape Metrics and Biodiver sity. Living Review in Landscape Research 5(3). Weibull, A.C., Ostman, O. and Granqvist, A. (2003). Species richness in agroecosystems: the effect of landscape, habitat and farm management. Biodiversity Conservation 12: 1335 1355. Westhoek, H., Van Zeijst H., Witmer, M., van den Berg, M., Overmars, K., ven der Esch, S. and van der Bilt, W. (2012) Greening the CAP An analysis of the Policy 2014 2020. PBL Netherlands Environmental As sessment Agency Wiens, A.J. (1992) What is landscape ecology, really? Landscape Ecology 7(3): 149 150
86 BIOGRAPHICAL SKETCH program of the Food and Resource Economics Department, in the framework of the inter Atlantic ATLANTIS program. Before graduating with a Master of Science in summer 2013, he obtained his Bachelor of Science at Ghent University, in spring 2011. During his studies, Louis was elected as student represe ntative of the ATLANTIS program. H e also participated in volunteer work for the inclusion of minorities at Ghent University.