Landscape Ecology

, Volume 25, Issue 6, pp 825–838

Designing agricultural landscapes for natural pest control: a transdisciplinary approach in the Hoeksche Waard (The Netherlands)


    • Wageningen University and Research Centre, Alterra
  • Willemien Geertsema
    • Wageningen University and Research Centre, Alterra
  • Walter K. R. E. van Wingerden
Landscape Ecology in Practice

DOI: 10.1007/s10980-010-9489-7

Cite this article as:
Steingröver, E.G., Geertsema, W. & van Wingerden, W.K.R.E. Landscape Ecol (2010) 25: 825. doi:10.1007/s10980-010-9489-7


The green–blue network of semi-natural non-crop landscape elements in agricultural landscapes has the potential to enhance natural pest control by providing various resources for the survival of beneficial insects that suppress crop pests. A study was done in the Hoeksche Waard to explore how generic scientific knowledge about the relationship between the spatial structure of the green–blue network and enhancement of natural pest control can be applied by stakeholders. The Hoeksche Waard is an agricultural area in the Netherlands, characterized by arable fields and an extensive network of dikes, creeks, ditches and field margins. Together with stakeholders from the area the research team developed spatial norms and design rules for the design of a green–blue network that supports natural pest control. The stakeholders represented different interests in the area: farmers, nature and landscape conservationists, water managers, and local and regional politicians. Knowledge about the spatial relationship among beneficial insects, pests and landscape structure is incomplete. We conclude that to apply scientific knowledge about natural pest control and the role of green–blue networks to stakeholders so that they can apply it in landscape change, knowledge transfer has to be transparent, area specific, understandable, practical and incorporate local knowledge.


Field marginsSpatial normsGreen–blue networkStakeholdersDecision-makingScaleFunctional biodiversityEcosystem servicesPesticidesBoundary object


European policy promotes sustainable agricultural practices aimed at reducing nutrient and pesticide input into the soil and regional water system. For example, farmers and water boards will soon have to meet the stringent water quality standards laid down in the EU Water Framework Directive (European Union 2000). However, removing pesticides from surface water is expensive, and some of the costs must be borne by farmers. So it is in the farmers’ interests to apply agricultural practices that are less dependent upon pesticides yet do not increase the risk of pests damaging crops, causing crop yield loss.

An alternative to chemical pesticides is natural pest control; this exploits the potential of arthropod species in the agricultural landscape to suppress pest insect populations by predation or parasitism (Altieri and Letourneau 1982; Kromp 1999; Bianchi et al. 2006). Three strategies can be distinguished for measures that support natural pest control. The first is soil management: alternative tillage programmes (Kromp 1999) support beneficial soil-dwelling species. The second strategy, crop management, entails applying measures such as mixed cropping or intercropping to augment the suppression of pest insects (Perrin and Phillips 1978; Theunissen et al. 1995). The third strategy comprises actions related to the landscape structure, by exploiting the fact that semi-natural landscape elements provide resources that support beneficial species that suppress pest species (Landis et al. 2000; Wäckers 2004; Bianchi et al. 2006). In this paper we focus on this third strategy.

We consider the configuration of all semi-natural landscape elements to be a functionally coherent ecological network, called the green–blue network. In agricultural landscapes, this green–blue network enhances predation and parasitism and may reduce crop damage by insect pests (Thies and Tscharntke 1999). The network provides several resources for the survival of natural enemies. It provides food for insects that need nectar or pollen (Wäckers 2004), winter habitat (Varchola and Dunn 2001), an attractive microclimate (compared to the extremes in the cropped area) (Langellotto and Denno 2004) and alternative prey after crop harvest (Altieri and Whitcomb 1979).

Measures to improve the green–blue network for the function of pest control at an adequate level should be based on the relationship between physical characteristics and functioning. Some studies have found a correlation between the spatial structure of green–blue networks and the effectiveness of the natural pest control (review in Bianchi et al. 2006). Other studies have described the relations between the vegetation composition and structure and the presence of beneficial insect species (Thomas et al. 1991; Wratten 1992; Landis et al. 2000). The benefits of native plants are mentioned by Isaacs et al. (2009). Vegetation composition and structure are influenced by management actions such as mowing and grazing (Hobbs and Huenneke 1992; Bakker 2005). Our work in the case study area focused on vegetation composition and structure (results not discussed in the present paper, Geertsema et al. 2004, 2006), as well as on the spatial structure of green–blue networks.

The question we addressed was how to transform scientific knowledge so that it can be applied by farmers and other stakeholders wishing to develop the landscape in such a way that it supports natural pest control. This question arose in the Hoeksche Waard, an area in the southwest of the Netherlands. Here, the extensive network of dikes, creeks and ditch banks makes the area very suitable for the implementation of natural pest control. The notion that sustainable agricultural practices are the future of the Hoeksche Waard was supported not only by the local farmers, but also by the regional water board, nature organizations, and local governments. The stakeholders had different views about which measures were adequate to realize sustainable agricultural practices. In general, the stakeholders were keen to use the latest scientific knowledge on natural pest control. This is only possible if scientific knowledge is translated into applicable spatial norms and design rules (Nassauer and Opdam 2008).

We opted to develop a stakeholder-oriented approach. Involving stakeholders in the design process has the advantages of creating local support, incorporating local knowledge, and connecting and binding stakeholders to the same vision. But to achieve cooperation among stakeholders, the different interests and backgrounds have to meet each other around a shared concept that can function as a boundary object. Boundary objects are defined as collaborative efforts/outputs that “are both adaptable to different viewpoints and robust enough to maintain identity across them” (Cash et al. 2003; Turnhout et al. 2007). We used the concept of green–blue networks as a boundary object to unite stakeholders with different interests in a shared vision of the future.

The aim of this paper is to describe, and reflect on how, we designed a green–blue network for natural pest control in an agricultural area. We worked jointly with local stakeholders and used the latest scientific knowledge about natural pest control. Our approach was to answer the following questions:
  1. 1.

    Can we find enough information about the quantitative relation between green–blue networks and natural pest control so that stakeholders can take adequate measures?

  2. 2.

    How can generic scientific knowledge about natural pest control be applied to the Hoeksche Waard, taking into account region-specific characteristics such as locally grown crops and their pests and natural enemies, and the typical characteristics of the green–blue network in the study area?

  3. 3.

    How can the design benefit from stakeholder engagement and local knowledge? How can stakeholders determine the actual state of the green–blue network in the Hoeksche Waard with reference to the desired functioning?


We limited our study to aboveground arthropod pest species and arthropod natural enemy species in the dominant cropping system in the case study area. The focus was mainly on the spatial structure of the green–blue network and less on the habitat quality required for effective natural pest control. We will report about the effect of the design method on the stakeholder decision-making process in a forthcoming paper (Opdam and Steingröver In prep.)

The case

Study area

The Hoeksche Waard lies south of the city of Rotterdam in the Netherlands and is bordered by major watercourses (Fig. 1). It is ca. 26,500 ha in area. The majority of its 85,700 inhabitants live in 14 small villages. The landscape consists of polders, reclaimed gradually after the Saint Elisabeth flood in 1421, the most recent reclamation being 200 years ago. The soils are mainly light clay and sandy clay. The polders are separated from each other by a network of dikes. The landscape is highly valued for its characteristic structure of polders, dikes and network of creeks; its quietness and openness compared with the dynamics of the urban area of Rotterdam are also highly valued. The area has the legal status of “national landscape”, to protect and conserve its characteristic landscape structure.
Fig. 1

Overview of the study area, inset: location of the Hoeksche Waard area in the Netherlands

The basis of this characteristic landscape is agricultural production: most of the polders (63% of the total area) are under arable crops, except for one polder in the central eastern part, which is predominantly used for dairy farming and so is under grassland. Most of the farmers apply conventional intensive farming, but there are some organic farmers in the area. The dominant crops in the arable farming system are potatoes, sugar beet, winter wheat and cabbage. The most common rotation is potatoes, sugar beet, wheat and fallow land or another crop. The most obnoxious pest species in this rotation are aphids and thrips. The most obnoxious pest species in cabbage crops are butterflies, moths and snails. In the southeastern part of the area, scientists and farmers (most of them are conventional farmers) have been experimenting with field margins to support natural pest control since 2004 in the so-called FAB-experiment (functional agro biodiversity). The participating farms in the experiments comprise an area of 400 ha (Van Alebeek et al. 2006b).

Semi-natural vegetation occurs in different types of landscape elements in the Hoeksche Waard: fine elements (maximum width 5 m) and robust elements (width more than 5 m). The fine elements consist of field margins, ditch banks and the verges of secondary roads, forming a network that entwines around the arable fields. The robust elements consist of dikes, creeks—sometimes bordered by patchy wetlands—forest patches and verges of the main roads. Most of the main roads are on the dikes. A characteristic of the area is that many of the dikes are flanked by lines of trees. The vegetation of the green–blue network tends to be rather species-poor, except in a small proportion of the dikes, creek banks, ditch banks and field margins (Photos 1, 2, 3).

Position of stakeholders

Stakeholders often have interests in ecosystem services that function at different scales (Hein et al. 2006). In our case, the stakeholders not only have interests at different scales—from farm to the entire Hoeksche Waard—but their interests also differ—from farmers with private interests to councils with public interests. Although all stakeholders had a mutual interest in preserving the identity and the agricultural character of the area and were aiming at more sustainable agricultural practices, each group also had specific interests. The water board was mainly interested in the resulting improvement in water quality, and the nature organizations in increasing biodiversity. For farmers, however, natural pest control must be not only effective, but also cost-efficient by saving on the costs of chemical pesticides and of water purification.


The building blocks for our design method are a spatial concept as boundary object, spatial norms and design rules. They are applied together with stakeholders, using state-of-the-art scientific knowledge.
  1. 1.

    Our design method is based on the spatial concept of green–blue networks, a suitable concept for landscape planning as it (a) bridges the gap between landscape conservation and development, (b) facilitates stakeholder decision-making, and (c) is the multifunctional carrier of landscape services (Opdam et al. 2006).

  2. 2.

    We define spatial norms as minimum measures of the spatial dimensions of the elements of the network which are needed to achieve a desired target (in our case, a certain level of natural pest control).

  3. 3.

    Design rules are guidelines for stakeholders when applying spatial norms in concrete situations. The research group developed the design rules, but for their actual application the involvement of the stakeholders was essential.


Making uncertainty explicit

We started by compiling the available information on the key processes and the accompanying spatial conditions for the functioning of natural pest control. This information was grouped into three types of sources: published field data on the Hoeksche Waard or comparable agricultural areas, data from simulation modelling, and expert knowledge from entomologists with field experience. These three sources have different degrees of uncertainty. In order to be suitable for the development of norms for this study, the data that we used had to be applicable to comparable crops, pests or natural enemies. We assumed that the least uncertain was the knowledge consisting of published field data. The knowledge derived from simulation models was more uncertain because it was based on more assumptions than the knowledge derived from field data. In contrast to these data sources, expert knowledge is not published and its replicability is questionable; therefore we assumed it was the most uncertain source of information.

This information and the uncertainty in it was communicated and discussed with the stakeholders in workshops:
  1. 1.
    Diagnosis and design workshop. Present were the research team, individual farmers, nature conservationists and representatives of farmers associations and nature organisations (ca 15 people). Uncertainty of the spatial norms (see codes in Tables 1 and 2) was communicated by colour coding. Using design rules the stakeholders diagnosed the present situation and designed the green–blue network on the regional and farm scale (Figs. 3 and 4) (Photo 4).
    Table 1

    Summary of the spatial norms for the density of fine elements, dependent on the presence or absence of robust elements in the landscape and also indicating the source (E expert knowledge, M simulation model, F published field data)


    Robust elements present

    Robust elements present

    Robust elements absent

    Distance from robust element

    0–75 m (E)

    75–1,000 m (F)

    >1,000 m (F)


    Active movement

    Passive movement

    No influence

    Additional fine elements needed?


    Yes (every 150 m) (M)

    Yes (every 100 m) (F)

    Table 2

    Summary of the spatial norms for the size of fine and robust elements and indication of the source (E expert knowledge, M simulation model, F published field data)




    Robust elements

    >25 m wide (E)

    >1 ha (M)

    Fine elements

    >3.5 m wide (F)


    n.a. not applicable in this study

  2. 2.

    Communication and uncertainty workshop. Present were the research team, the stakeholders that attended the first workshop, regional and local policy makers (e.g. municipalities, water board) and invited scientific experts on entomology (ca 25 people). The results of the first workshop were presented, the experts commented on the spatial concept and the uncertainties involved, and the local policy makers discussed the consequences for implementation.


The green–blue network as provider of natural pest control

For natural pest control to be effective, the populations of natural enemies must be able to reach the population of pests at the right time and at the right place to prevent the pest population exceeding a critical density above which there is significant loss of crop yield. We need to know which factors influence the number of natural enemies reaching the pests, and which factors can be adjusted by means of the design and management of the green–blue network.

Many studies have demonstrated the correlation between the presence of semi-natural habitats within agricultural landscapes and the diversity and density of natural enemies (see Bianchi et al. 2006 for details). This correlation comes about because as well as needing arable fields, natural enemies also need semi-natural vegetation for population survival. They need this vegetation for several reasons:
  • Food: flowering plants are an important source of nectar and pollen for the adult stage of parasitic wasps or hoverflies (Bugg and Picket 1998; Nentwig 1998; Wäckers 2004; Winkler et al. 2006).

  • Shelter: permanent vegetation and the presence of certain structures (e.g. dead plant material) provide shelter for insects during the winter and against weather extremes (heat, rainfall) during the growing season (Duelli et al. 1990; Dennis and Fry 1992; Dyer and Landis 1996; Nentwig 1998).

  • Stable environment: the environmental conditions in green–blue networks are stable and continuous, compared to the arable fields, where there is annual turnover of the vegetation (Landis and Menalled 1998; Nentwig 1998).

These conditions are provided by several vegetation types, ranging from herbaceous, through ruderal to woody vegetation. So a variety of elements in the green–blue network potentially contribute to the survival of natural enemy populations, providing that the vegetation structure and composition afford food or shelter.

The spatio-temporal dynamics of natural enemies and pest species and their dependency on the green–blue network differ. This difference between the two species groups is the basis for the potential of natural pest control via the green–blue network. Many natural enemies actively move between green–blue elements and crops daily: they find shelter, nectar and pollen in the semi-natural vegetation in the green–blue elements, and prey and hosts in the crops. While most natural enemies use the green–blue networks daily, e.g. parasitoid wasps need to feed on nectar to obtain energy for flying and searching for hosts (Wanner et al. 2006), the pest species can find the food sources for their daily survival in the crops (herbivory on the crops).

Besides active movement, passive movement occurs when natural enemies or pest species are transported by wind (Chapman et al. 2004). In seasonal movement between overwintering habitat and reproduction habitat this plays an important role, augmenting the active movement over shorter distances (Corbett and Rosenheim 1996; Wissinger 1997). In spring, passive movement over large distances provides a basic density of natural enemies over large areas; spiders are valuable aerial immigrants into newly planted crops (Sunderland and Samu 2000). Robust semi-natural landscape elements are an especially important source of airborne insects (Grüebler et al. 2007).

Pest insects need alternative host plants after the crop is harvested and may find these plant species in the green–blue network. Some species survive in semi-natural vegetation in the green–blue networks during the winter season (Moeser and Vidal 2004; Roschewitz et al. 2005). However, in many studies, the positive relation between the green–blue network and abundance has not been very strong, or it has been more apparent at larger scales for pest insects (Thies et al. 2003). An explanation is that pest insects may be transported by wind over very large distances, over 400 km (Wolf et al. 1990). Thus, in contrast to beneficial insects, the distribution of pest insects is less sensitive to the spatial structure of the green–blue network in the landscape, at least at spatial scales larger than a few kilometres. Within the scale of a few kilometres, positive relationships between abundance and landscape structure are found for both pests and natural enemies (Roschewitz et al. 2005).

Figure 2 illustrates the relationship between robust elements and fine elements as sources of natural enemies. This is the basis for the norms for mesh size of the network, defined by the distance between fine elements.
Fig. 2

Illustration of the norm for the distance between fine elements when robust elements are present (top panel) or absent (central panel). Due to the presence of robust elements as a source for passive movement, the norm for maximum distance between fine elements is higher when robust element are present, than when they are absent

Spatial norms for the green–blue network

Spatial norms are derived from different data sources (see section ‘Making uncertainty explicit’). Spatial norms quantify the minimum requirements for different spatial characteristics of the green–blue networks and of the elements that make up the network. Quantitative information on the relation between spatial dimensions and the effectiveness of natural pest control was scarce. Therefore, we formulated the spatial norms in terms of minimum values. An overview of the spatial norms is given in Tables 1 and 2.

In order to fulfil the function of providing food and shelter for the natural enemies, the elements should be at least 3.5 m wide (Thomas 2000). The distance covered by active movement is limited and varies between species. Lady beetles can cross a distance of 30–100 m (Corbett 1998; Bianchi and Van der Werf 2003). Parasitoid wasps can cross distances of ca. 60 to ca. 250 m (Lavandero et al. 2005; Wanner et al. 2006).

Studies on the relation between the presence of surrounding non-crop habitats at different landscape scales and observed parasitism or predation abundance of natural enemies have often demonstrated that the effect of robust landscape elements (grassland, forest, fallow, etc.) extends to at least 1 km (Elliot et al. 1998; Sunderland and Samu 2000; Bianchi et al. 2005, 2008; Purtauf et al. 2005). We know of no published data on the quantitative relation between the size (area or width) of individual robust elements and their effect on natural pest control in the surrounding landscapes, but the experts consulted in the second workshop indicated that an area of 1 ha or a width of 25 m would be sufficient.

When robust elements are absent and fine elements are the only source of natural enemies, the major source of the beneficial insects for natural pest control is active movement. Since most movement occurs over distances shorter than 50 m, we assume that the distance between fine elements should not exceed 100 m. When robust elements are present, both active and passive movements are sources of natural enemies. Given that passive movement from robust elements is influential over at least 1 km, we propose that within a distance of 1 km from robust elements, the distance between fine elements as a source of active movement can be as large as 150 m.

Design rules

Based on the required vegetation composition and the spatial norms, a set of design rules was formulated that was applied by stakeholders at the scale of the Hoeksche Waard as well as at the scale of individual farms. The design rules were applied in the diagnosis and design workshop:
  1. a.

    Determine which green–blue elements may function as robust elements in natural pest control.

  2. b.

    Determine whether the qualitative and spatial characteristics of the robust elements meet the requirements for natural pest control. If necessary, formulate the adaptation measures.

  3. c.

    Determine which green–blue elements may function as fine elements in natural pest control.

  4. d.

    Determine whether the qualitative and spatial characteristics of the fine elements meet the requirements for natural pest control. If necessary, formulate the adaptation measures.


Application of the approach


The robust elements of the green–blue network in the Hoeksche Waard (dikes, creeks and road verges) are distributed over the whole area. Almost all (>99%) the arable land lies within a radius of 1 km from the robust elements. According to the stakeholders, most of the dikes are wide enough to meet the norm of a width of 25 m (Table 2), but some of the creek banks fail to meet this criterion. At present, the majority of the dikes and creek banks are traditionally managed (short vegetation due to frequent mowing) and make no significant contribution to natural pest control. To function effectively as sources of natural enemies, the management of the robust elements should be adjusted to favour pest control by providing conditions for food, shelter and a stable environment.

The fine elements of the green–blue network (ditch banks, road verges and field margins) are also distributed over the whole area. Applying the norms, the width of the separate elements is usually sufficient, but the density is too low. To function for natural pest control, the network of fine elements has to be extended. As was the case with the robust elements, the majority of the road verges, ditch banks and field margins are traditionally managed and so, once again, to effectively support natural enemies, the management of the fine elements should be adjusted to favour pest control by providing suitable vegetation structure and composition.

Efficiently achieving natural pest control

The stakeholders involved in the project indicated that, ideally, the arable area in the Hoeksche Waard that is effectively protected by natural pest control comprise all the arable land. The Hoeksche Waard stakeholders aspired to move forward from the present situation to this ideal situation as efficiently and rapidly as possible. They did not opt for a fixed, quantified target at a certain point in time (e.g. 25% arable land protected by natural pest control in 5 years).

To enable the stakeholders to achieve this, different strategies were developed, based either on spatial effectiveness, through a focus on the green–blue network, or on organizational networks (stakeholders with shared interests). The resulting four strategies were not mutually exclusive:
  1. 1.

    “Robust elements first” aims at ensuring that all dikes and creeks function as sources of beneficial insects for natural pest control (spatial strategy);

  2. 2.

    “Priority areas” aims at achieving the ideal situation as soon as possible in part of the Hoeksche Waard, using the experience gained from this to ensure that implementation in the rest of the Hoeksche Waard will be more efficient (organizational strategy);

  3. 3.

    “Combine mutual interests” aims at uniting stakeholders with similar interests (organizational strategy);

  4. 4.

    “Expand present fine networks” aims at enlarging and increasing the density of current networks of fine elements (spatial strategy).

Figure 3a illustrates strategy 1. It shows the consequences of adjusting management of all robust elements (and after adjusting the width of elements when necessary). Strategy 4 is illustrated in Fig. 3b. It shows the consequences of adjusting management (and the width of elements when necessary) of the current robust elements and of the ditch banks. Figure 4 shows the application of the spatial norms at farm level in three steps, as designed by one of the farmers in the diagnosis and design workshop. The first step shows the consequences of adjusting management and, where necessary, the width of the robust elements (Fig. 4a), the second step shows the additional consequences of adjusting the fine elements (Fig. 4b), and the third step shows the additional consequences of adding new fine elements needed to meet the norms for the density of the green–blue network (Fig. 4c). The consequences of the spatial strategies 1 and 4 are quantified in Table 3.
Fig. 3

Potential of Hoeksche Waard area for natural pest control, depending on which existing landscape elements are managed. a all robust elements managed for natural pest control (the arable land within 75 m from robust elements is not shown because not clear at this scale); b all robust elements and ditch banks managed for natural pest control
Fig. 4

Potential for natural pest control on an individual farm through adapting management. a robust elements; b robust and existing fine elements; c existing robust and fine elements and creation of new fine elements

Table 3

Planned investments in green–blue networks and projected area under desired natural pest control


Robust first (Fig. 3a)

Robust + fine (Fig. 3b)

Area level

Investments in green blue networks

 Management of robust elements (km)



 Management of existing fine elements (km)

196.7 (existing field margins) and 2181.6 ditch banks

Results for natural pest control

 Arable land within active movement distance (ha)



 Arable land within passive movement distance (ha)



 Arable land outside passive movement distance (ha)




Robust elements (Fig. 4a)

Robust + fine elements (Fig. 4b)

Robust + fine + new elements (Fig. 4c)

Farm level

Investments in green–blue networks

 Management of robust elements (km)




 Management of existing fine elements (km)




 Creating new elements (km)




Results for natural pest control (total farmed area = 130 ha)

 Arable land within active movement distance (ha)




 Arable land within passive movement distance (ha)




aAssuming both dikes and creek banks are sufficiently wide
Photo 1

Brussels sprouts, one of the typical crops
Photo 2

Creek with reed vegetation bordering arable fields
Photo 3

Dike flanked by willow trees adjacent to arable fields
Photo 4

Workshop with stakeholders from the Hoeksche Waard

Choice of strategies

At this stage of the process, a third workshop was organized: the strategies workshop. Present were the research team, individual farmers and representatives of farmers associations, nature organizations, the water board and municipalities (ca 15 people). The stakeholders discussed the presented strategies and preferred to implement them all. They decided to start with strategy 2 (priority areas), but also take advantage of opportunities outside the priority areas (combination of strategies 1, 3 and 4). Two priority areas were chosen:
  • Priority area 1: Opportunities through the presence of the FAB (Functional Agro Biodiversity) experiments on farms in this area (Van Alebeek et al. 2006a, b) and through the presence of high-quality robust elements (nature development areas).

  • Priority area 2: Opportunities through shared interests because of the presence of nature areas and combination with recreation. At the same time, area 2 could profit from lessons learned in area 1.

Opportunities outside the priority areas were the presence of field margins from the arable margin project, and connections with nature areas. The argument for choosing this combination of strategies was that low investments would already lead to more of the area being under natural pest control and would boost the motivation of stakeholders to participate in developing the green–blue network for natural pest control.


The main question of this paper was how to transform scientific knowledge about natural pest control and the role of green–blue networks to stakeholders in such a way that they can apply it in landscape change. Cash et al. 2003 argue that to be effective knowledge transfer needs to be credible, salient and legitimate. We conclude that knowledge transfer has to be transparent and area specific. And in order to be applied by stakeholders, it has to be understandable, practical and incorporate local knowledge. Knowledge transfer was transparent by communicating the sources of the scientific knowledge, the uncertainty of the knowledge and the gaps in the knowledge to the stakeholders. The credibility of knowledge transfer was also ensured by organizing the communication and uncertainty workshop and discussing the derivation of the norms with the stakeholders.

We conclude that the knowledge transfer was salient as the developed norms were area specific, taking into account the spatial characteristics of specific elements of the green–blue network in the Hoeksche Waard. Also, the diagnosis and design workshop showed that knowledge transfer was understandable and practical as the stakeholders familiarized themselves easily with the relation between the green–blue network and natural pest control and incorporated local knowledge into their design for natural pest control. The legitimacy of the knowledge transfer (Cash et al. 2003) was addressed by recognizing and discussing the interests of different groups of stakeholders.

Is there enough information about the quantitative relation between green–blue networks and natural pest control so that stakeholders can take adequate measures?

The ultimate goal of the design of the green–blue network was the development of sustainable farming systems that do not depend on chemical pesticides and that are economically viable. That also means that natural pest control has to be so effective that crop yields are not affected by pest damage. The data from field trials are promising. Recent results from the ongoing FAB experiment in the Hoeksche Waard on summer wheat and potatoes show that in most years aphids are kept below the damage threshold without additional control measures (Wiersema and Van Alebeek 2006; Van Alebeek et al. 2006b). FAB experiments have now been set up in other parts of the Netherlands.

Field studies such as the FAB experiments are an important source of information in a common learning process of researchers and farmers. So far, most papers about landscape structure and natural pest control focus on the correlation between landscape features and the diversity or density of natural enemies; some focus on the density of pests (e.g. Bianchi et al. 2006, 2008; Kremen and Chaplin-Kramer 2007), but only a few have investigated yield loss. Yet this issue will determine the farmers’ attitude towards natural pest control. So, integration of structural, functional and valuation aspects are imperative for the successful planning of ecosystem services like natural pest control (Termorshuizen and Opdam 2009).

How can generic scientific knowledge about natural pest control be applied to the Hoeksche Waard: managing uncertainty

Thies and Tscharntke (1999) have argued that we do not know enough to be able to generalize about the relation between landscape structure and populations of natural enemies. Their main argument is that different organisms probably respond very differently to landscape structure. Bianchi et al. (2006) also emphasize that more research is needed before knowledge about the landscape—natural enemies system can be generalized. Though we agree that there is still a large gap of knowledge to close and norms are heuristic, we believe that the crucial question seems to be who determines whether and on what grounds norms are good enough to be applied in the field. Can the scientist decide on his own, using scientific criteria, or can the decision be taken jointly with the stakeholder? If scientists are transparent about the uncertainties in norms, stakeholders will be able to decide whether it is worth them taking the risk. A dialogue between scientists and stakeholders may yield the pros and cons of applying the norms in the specific situation of the stakeholders and strengthen their resolve to use current scientific knowledge. Or, as Cash et al. (2003) state: linking knowledge to action requires open channels between experts and decision makers. So, scientists have two options: to continue doing research and ignore stakeholders’ requests for application, or to translate current knowledge with precaution, working closely with stakeholders, in order to make them aware of the uncertainty. Classifying the spatial norms according to their uncertainty is important to stakeholders, as it relates to the risks involved. Learning by monitoring is crucial element to this option.

The spatial norms developed for the Hoeksche Waard are not final; they reflect knowledge at this point in time and they should be adjusted where and when necessary.

Translation to different landscapes

Although the norms were developed for the Hoeksche Waard, green–blue networks can also provide natural pest control in other landscapes. The combination of robust and fine elements in a landscape is a feature common to many landscapes in northwestern Europe, parts of the US and Canada. Isaacs et al. (2009) state that the success of insect conservation on farmland depends on landscape context, with the probability of success being highest in landscapes of moderate complexity. The density of the network may differ, but the concept of using both robust and fine landscape elements is applicable to other landscapes as well. The network of linear landscapes that is typical of the Hoeksche Waard may not be typical of other landscapes. In other landscapes, robust elements may be more patchy and therefore norms that relate to the patchiness will be more appropriate (e.g. patch area and percentage of the landscape covered by patches). Translation to other landscapes requires the presence of similar crops, pests and enemies as well, though this will not be a problem in a northwestern European context.

Recent literature on planning ecosystem services suggests that implementation in regional contexts should be based on the characteristics of the ecosystems of interest and on the decision-making context for which the planning has been mobilized (Fisher et al. 2008), and that regional approaches may offer a more constructive way forward than more generic approaches (Anderson et al. 2009).

How can the diagnosis and the design benefit from stakeholder engagement and local knowledge?

The incorporation of local knowledge played an essential part in the design process by adding area specific information about:
  • the current status of the green–blue network which led to a distinction between ecological management aimed at conservation of plants and butterflies, and management aimed at natural pest control, which required a larger variety of plants and insects. This improved the diagnosis of the current situation.

  • the elements of the green–blue network. The farmers added the embankments of the high-speed train line to the robust elements that could be used and managed for natural pest control. They also suggested GPS during ploughing to create temporary mid-field islands of semi-natural vegetation when fields are too wide. This improved the design.

  • the regional social structures. This was crucial to formulate spatial and organizational strategies, because the stakeholders recognized that implementation would benefit greatly from coalition building. This improved the chances of successful implementation.

A multi-stakeholder approach

We argue that the implementation of natural pest control to reach sustainable agricultural practices is not only of concern to farmers, but also to other stakeholders. Implementation leads to improved water quality which is important to the local water boards and communities. It improves the attractiveness of the landscape which increases recreational use and the values of houses. For each group of stakeholders, the balance between costs and benefits will be crucial in the decision to participate in plans for natural pest control. A key factor in this balance is the compatibility of natural pest control with other functions. A separate cost-benefit analysis of the investments and profits from green–blue networks in the Hoeksche Waard concluded that the combination of functions is often possible and financially beneficial (VROM 2007). For such networks to be successfully implemented, farmers should be remunerated for providing environmental services from agricultural land (Dobbs and Pretty 2008), as in the case of compensating cattle rangers for providing water storage and nutrient retention (Bohlen et al. 2009).

The stakeholders’ willingness to participate in the development of the green–blue network for natural pest control offers opportunities for coalition-building between private and public stakeholders. In other areas, stakeholders may be less cooperative and this may have consequences for the strategies that are applicable. The implementation must also been seen in the light of the socio-economic context of the area involved. Developments in agricultural and environmental policy and legislation at national and European scales influence the acceptance of natural pest control. Although uncertain, the trend in agricultural and environmental policy and legislation in the last decade has been in favour of more environmentally friendly farming systems. For instance, the CAP (Common Agricultural Practice), the system of European Union agricultural subsidies and programmes is changing from crop subsidies to “licence to produce”. This means among other things the stimulation of crop production systems with less negative impacts on the environment. The reduction of pesticides fits in here. A recent study on farmlands across Europe showed the negative impacts of pesticides on biodiversity and on the biodiversity-based ecosystem service of natural pest control (Geiger et al. 2010), giving arguments for the development of sustainable farming systems. We hope that the approach presented in our paper gives further support for the development of sustainable farming systems.


This research was financed by the province of South-Holland and by the Dutch Ministry of Housing, Spatial Planning and the Environment (VROM). Joy Burrough advised on the English.

Copyright information

© Springer Science+Business Media B.V. 2010