SEARCH

SEARCH BY CITATION

Keywords:

  • agri-environmental schemes;
  • bees;
  • conventional farming;
  • hedgerows;
  • hoverflies;
  • intensive grasslands;
  • network structure;
  • pollen limitation;
  • specialization asymmetry;
  • Trifolium sp.

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information

1. Pollination interactions comprise a network of connections between flowers and insect visitors. They are crucial for reproductive success in many angiosperms but are threatened by intensive agricultural practices. Although less intensive approaches, including organic farming, could improve farmland biodiversity, it is not clear whether or not these approaches enhance wild plant pollination and the stability of insect–flower interaction networks.

2. We investigated the effects of organic vs. conventional farming on insect–flower interaction network size and structure, bee and hoverfly diversity, and pollination in 10 pairs of organic and conventional dairy farms in the Republic of Ireland.

3. We found that insect–flower interaction networks on organic farms were larger and more asymmetrically structured than networks on conventional farms. Overall, however, networks contained fewer taxa and niche overlap and plant/animal ratios were relatively low compared with previously documented insect–flower interaction networks. Organic farms did attract higher numbers of bees partly because of higher floral abundances (mainly Trifolium sp.). Hoverfly evenness was greater in organic farms but neither abundance, richness nor evenness was related to floral abundance, suggesting organic farms provide additional resources for hoverflies. Pollination of Crataegus monogyna hawthorn was higher on organic farms, although pollen deposition was limited.

4.Synthesis and applications. Organic dairy farming can increase the size and alter the structure of insect–flower interaction networks. However, network stability was not improved and all networks (organic and conventional) were vulnerable because of their small size, low niche overlap and low plant/animal ratios. Nonetheless, organic farming provided more flowers that attracted more flower visitors and improved pollination of C. monogyna. We suggest that strategic management of important flowers for pollinators in hedgerows and pastures should be endorsed in agri-environmental schemes. Sowing Trifolium spp., and allowing these plants to flower, could benefit bees, but more research into hoverfly ecology is necessary before realistic conservation recommendations can be made for this group. We conclude that organic farming, although not the solution in its present form, can benefit insect biodiversity, insect–flower interaction networks and insect-mediated pollination.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information

Many wild plant populations and agricultural crops are dependent on pollination services provided by wild pollinators, such as bees, hoverflies and butterflies (Free 1993; Biesmeijer et al. 2006). However, wild pollinators have been negatively affected by agricultural intensification and numerous pollinating taxa are now in decline in Europe and North America (Steffan-Dewenter, Potts & Packer 2005). Bees are considered the most effective pollinators of wild plants but hoverflies have also been shown to provide significant pollination services (Gyan & Woodell 1987a; Sugiura 1996; Vance, Bernhardt & Edens 2004). Few studies have investigated the effects of land-use change on hoverflies (e.g. Jauker et al. 2009; Meyer, Jauker & Steffan-Dewenter 2009).

In theory, organic farming (European Union Regulation 2092/91/EEC) should increase biodiversity because it is less intensive than conventional farming (Bengtsson, Ahnstrom & Weibull 2005). Beneficial effects of organic farming on plant and pollinator diversity have been reported for arable systems (Bengtsson, Ahnstrom & Weibull 2005; Morandin & Winston 2005), but few studies have compared the biodiversity of organically and conventionally managed grasslands (but see Potts et al. 2009). Also, the majority of studies to date on grassland biodiversity relate to species-rich or ‘high biological value’ grasslands. This is unfortunate given that intensive management is dominant in most lowland grasslands, representing millions of hectares in Europe (Plantureux, Peeters & McCracken 2005). In the Republic of Ireland, more than 80% of agricultural land is permanent grassland, mostly intensively managed (Department of Agriculture, 2009). Intensive grasslands receive high fertilizer application rates and frequent intensive defoliation which optimizes harvested forage quality but results in degraded species pools and structurally homogenous swards (Vickery et al. 2001). Organically managed grasslands would be considered less intensive than conventional grasslands owing to: prohibition of chemical fertilizers and pesticides; encouragement of nitrogen fixation (through planting of Trifolium spp. and reduction in manure and slurry fertilizers inputs) and lower grazing intensities. However, both organic and conventional grasslands are still intensive grasslands because they receive anthropogenic disturbance above and beyond the capacity for most disturbance/nutrient sensitive plant species to survive.

This study focused on the role of organically vs. conventionally managed dairy grasslands in supporting wild flower-visitor communities and wild plant pollination. We investigated whether farm management influences insect–flower interaction network size and structure. Insect–flower interaction networks consist of interactions between communities of insects and plants that generally result in mutualistic benefits. In addition to traditional parameters for assessing impacts on biological communities, such as species abundance, richness and diversity indices, network analysis can be used to detect underlying changes to community structure, species interactions and ecosystem function (Tylianakis, Tscharntke & Lewis 2007). Interest in the structure of insect–flower interaction networks has grown in recent years because they allow characterization of the stability of pollination systems (Vázquez et al. 2009). We use this methodology to gain understanding of the pollination systems that persist in disturbed grassland habitats.

It is becoming clear that bees and hoverflies respond to habitat change in different ways (Kleijn & van Langevelde 2006; Jauker et al. 2009) and conservation actions that encourage bees may not necessarily favour hoverflies. It is not known how both bee and hoverfly assemblages are influenced by farming system in intensive grasslands. Therefore, our second hypothesis was that organic farming positively affects bee and hoverfly abundance, richness and evenness. Our study focused on the farm scale and within the local scale of the farm by comparing field centres with edges.

For plants, inadequate pollination services can lead to pollen limitation which can result in reduced seed/fruit set (Knight et al. 2005). For this reason, it is not enough to quantify insect assemblages and their flower visitation rates, without exploring actual plant reproductive success. Therefore, our third hypothesis was that farm management influences pollination success of a common hedgerow species. This is the first study to investigate relationships between insect–flower interaction networks, insect diversity and pollination in relation to farming system.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information

Study sites

Ten matched pairs of organic (managed according to the European Union Regulation 2092/91/EEC) and conventional (not managed according to organic regulations) dairy farms in lowland permanent grassland (not ploughed or reseeded for at least 8 years) in the Republic of Ireland were selected (see Table S1, Supporting information). Organic farms were certified for 11·5 years on average (range: 7–20 years), following a 2-year conversion period. Pairs were matched on geology, soil type and climatic similarity. Farms within a pair were 1–4 km apart, in central and southern Ireland (see Fig. S1, Supporting information), which is characterized by well-drained, fertile soils used predominantly for dairy/drystock farming. Average Livestock Units per hectare on organic farms were 1·5, compared with 2·5 on conventional farms.

Surveys of flower visitors and flowering plants

Two fields from each farm were surveyed three times each between May and July 2009 using standard line transects (Pollard 1977; Westphal et al. 2008). Transects 100 × 2 m were walked slowly (10–15 m min−1) along the edge (stock-proof hedgerows) and in the centre (30 m from the edge) of every field (12 transects in total per site). All bees (Hymenoptera: Apidae), hoverflies (Diptera: Syrphidae) and butterflies (Lepidoptera) observed within transects were recorded, together with each individual’s activity (feeding or not-feeding). If feeding, the respective flower species was noted. In addition to these obligate flower visitors, there may be other facultative flower-visiting taxa which act as pollinators in the study sites (including beetles, flies, etc.). However, we focused on bees and hoverflies because they are the most important and, often, the most effective pollinators of wild plants (Gyan & Woodell 1987a; Sugiura 1996; Vance, Bernhardt & Edens 2004). Specimens that could not be identified on the wing were caught, frozen and identified in the laboratory. Bumblebees (Bombus spp.), honeybees Apis mellifera, butterflies and solitary bees were identified to species and hoverflies to genus (using: Westrich 1989; Stubbs 2002; Shackleton, Nash & Lewington 2004; Edwards & Jenner 2005) (except for Bombus terrestris and Bombus lucorum, which were aggregated because of difficulties in distinguishing between them and other cryptic species using morphological methods; Murray et al. 2008; Wolf, Rohde & Moritz 2010). Both farms within a pair were visited on the same day and timing of visits within a pair, during the three survey periods, were alternated between morning and afternoon so as to reduce temporal biases. Surveys were conducted between 10:00 and 17:00 on days when ambient temperature was 14 °C or above with no rain or strong winds. Air temperature (°C), wind speed (Beaufort scale) and percentage cloud cover were noted for each transect.

Data on floral abundance and species richness were recorded in every transect in each survey. For every herbaceous flowering plant (excluding grasses, sedges and rushes), species identity (using Stace 2010) and the number of flowering units was estimated in 10 quadrats per transect (two 1 m2 quadrats every 20 m along transect). Flowering units were defined as an individual flower or compact inflorescence such that a medium-sized bee has to fly rather than walk between to reach another flowering unit (Dicks, Corbet & Pywell 2002).

Floral unit abundance and abundance, richness and evenness (Shannon Evenness) of bee species (bumblebees and solitary bees combined), hoverfly genera and butterfly species were calculated at the farm level, where edge and, separately, centre transects from three survey periods and two fields surveyed per site were pooled. Species accumulation curves were plotted to ensure the asymptote was reached before evenness was calculated. Butterflies were too rare at sites to be analysed but were included in interaction networks (see next).

Network analysis

A data matrix was constructed for each organic and conventional farm containing the total number of visits observed for each insect–flower interaction over the entire study period. Each matrix was illustrated by a quantitative interaction web (bipartite visitation graph), where a line between two vertices represents an interaction between an insect and a plant species. The interaction webs were then characterized by qualitative and quantitative network parameters using the ‘networklevel’ command in the bipartite package (Dormann et al. 2009) in R (R Development Core Team 2007). Quantitative parameters weight interactions according to their frequency which makes them less sensitive to sampling intensity and network size. Qualitative parameters included: number of plant and insect taxa in each network and the number of insect–flower interactions. The quantitative parameters calculated were: (i) Generality (weighted mean number of plant species per visitor species). (ii) Vulnerability (weighted mean number of visitors per plant species). (iii) Quantitative connectance (the realized proportion of all possible interactions weighted by the quantitative visitation rate of each species; Bersier, Banasek-Richter & Cattin 2002; Kaiser-Bunbury, Memmott & Müller 2009). (iv) Specialization asymmetry di′ (quantification of the imbalance between the interaction strengths of a species pair (Blüthgen et al. 2007; Dormann et al. 2009) in which singleton species were omitted (Dormann et al. 2009) and the effect of variation in observation totals was removed (Blüthgen et al. 2007); with 0 indicating high symmetry, and values close to 1 or −1 indicating high asymmetry in terms of specialization by insects or plants, respectively). (v) Niche overlap (mean similarity in interaction pattern between species of the same trophic level; values near 0 indicate no common use of niches, whereas 1 indicates perfect niche overlap; Dormann et al. 2009).

Pollination

Pollination was measured on all organic and conventional farms using Crataegus monogyna Jacq., hawthorn, as a model species. Crataegus monogyna was chosen because it is visited by bees, hoverflies and butterflies; it is self-incompatible (Grime, Hodgson & Hunt 2007); its flowers were located in field edges so disturbance by livestock was minimized; and it was the only suitable species naturally abundant at all study sites. We verified the reliance of C. monogyna on cross pollination: when insects were prevented from visiting 200 flowers using bridal veil netting, only four flowers (2%) produced fruit.

To assess natural levels of pollination, at each site 20 flowers (one per plant) in bud were marked and subsequently examined when fruits were beginning to form. Pollination was determined to have occurred if the hypanthium/ovary had swollen to form an immature fruit (i.e. the hypanthium had not withered). Fruit set was calculated as the ratio of pollinated vs. unpollinated flowers for each site.

Pollen limitation was investigated by conducting supplemental hand-pollination experiments on an additional 20 flowers per site. Flowers were marked whilst in bud, and once opened were supplementally hand pollinated using the stamens from flowers of different C. monogyna individuals along the same hedgerow. Before and after supplemental pollination, flowers were accessible to insects. Fruit set of experimental plants was determined 6 weeks after hand-pollination.

Statistical analysis

Network parameters and fruit set of open/supplementally pollinated flowers were investigated for differences between farming system (organic/conventional), whereas abundance/richness/evenness of bee/hoverfly taxa and floral abundance were tested for differences between farming system and position in field (edge/centre). Network parameters and abundance/richness/evenness of bee/hoverfly taxa and floral abundance were analysed using Linear Mixed Effects Models and fruit set of open/supplementally pollinated flowers was analysed using Binomial GLMM (cbind function). We accounted for spatial autocorrelation by including random terms: farm pair (1–10) and, when applicable, location-within-field (edge/centre). Plant species number (in networks), generality, floral abundance and hoverfly richness were square root transformed [√(x + 1)] and bee and hoverfly abundances were log transformed [ln(x + 1)] to achieve normality. In floral abundance and network models: farming system was included as a fixed effect. Floral abundance was included as a fixed effect, interacting with farming system, in the generality and ‘number of insect–flower interactions’ network models. Insect abundance (all bees, hoverflies and butterflies: feeding and non-feeding) was included as a fixed effect, interacting with farming system, in the vulnerability and specialization asymmetry network models. In the bee/hoverfly models: farming system, edge/centre, floral abundance, wind speed, air temperature, percentage cloud cover and bee/hoverfly abundance (see explanation in the following) were included as fixed effects as well as two-way and three-way interactions between farming system, edge/centre and floral abundance. In the richness and evenness models samples containing zero or one individual were removed and abundance was added as a fixed effect to ensure that richness and evenness models were independent from abundance models. Floral species richness was correlated with floral abundance (r = 0·257, = 0·001) and so was not analysed further. In the fruit set models: treatment (with four levels: Organic-Open, Conventional-Open, Organic-Supplemental and Conventional-Supplemental) was included as a fixed effect. Models were simplified by removing first non-significant interactions (> 0·05) and then any non-significant main effects (that were not constituent within a significant interaction). For model validation and model adequacy assessment, we followed Zuur et al. (2009). Statistical analysis was carried out using nlme (Pinheiro et al. 2009) and lme4 (Bates & Sarkar 2006) packages in R (R Development Core Team 2007).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information

In total, 504 bees (439 bumblebees, 57 honeybees and 8 solitary bees) were found from five bumblebee species, one honeybee species and three solitary bee species. A total of 832 hoverflies from 17 genera were found. Fewer butterfly individuals were recorded: 164 individuals of seven species. There were 39 species of flowering plant. Trifolium repens L. white clover was the most abundant flowering species in organic farms, with 3382 floral units in field centres, whereas only 177 T. repens floral units were counted in conventional centres. Bellis perennis L. common daisy was the most abundant flower in conventional farms with 568 floral units counted (compared with 1188 units in organic farms).

Flower visitors and farming system

Bee abundance and evenness were significantly higher on organic farms than conventional ones but bee species richness was not (Fig. 1, see Table S2, Supporting information). Independent of farming system, bee abundance and evenness were significantly lower in field centres compared with edges, and this pattern was more pronounced in conventional farms (farm type × edge/centre; Table S2, Supporting information, Fig. 1). Bee abundance was positively related to floral abundance, which was higher on organic farms (t9 = 5·52, < 0·001) and particularly high in organic field centres (t18 = 3·45, = 0·003) compared with conventional centres (Fig. S2, Supporting information). Bee abundance, richness and evenness were not related to environmental conditions.

image

Figure 1.  Geometric mean (+SE) abundance, richness and evenness of bees (bumblebees and solitary bee species) and hoverflies (genera) in edges and centres of organic and conventional fields. Significant differences in abundance/richness/evenness between organic and conventional or the edges and centres of all fields (independent of farming system) are illustrated by asterisks (*< 0·05; **< 0·01; ***< 0·001). Significant interactions between farm type and edge/centre are shown by ‘*i’ or ‘**i.’ Non-significant differences are not included.

Download figure to PowerPoint

Hoverfly abundance and richness were independent of farming system, but hoverfly evenness was significantly higher in organic compared with conventional farms (Table S2, Supporting information, Fig. 1). Also, hoverfly abundance, richness and evenness were lower in field centres compared to edges, regardless of farming system. Hoverfly evenness was negatively affected by wind speed (Table S2, Supporting information).

Insect–flower interaction networks and farming system

Organic and conventional insect–flower communities differed in size and composition. Although the number of insect species was similar between organic and conventional networks, significantly more plant species were visited within organic networks and, subsequently, generality was higher in the organic networks (Table 1, Fig. 2, see Figs S3–S11, Supporting information). Vulnerability and the number of insect–flower interactions were higher in organic networks but there were no interaction effects between farm type and flower abundance or insect abundance, respectively. Vulnerability was positively related to insect abundance. Insect species out-numbered plants in ratios of 1 : 1·2 for organic networks and 1 : 1·6 for conventional networks.

Table 1.   Mean (±SE) network parameter values for organic and conventional farms including significance levels, t-values and d.f. for explanatory variables obtained from Linear Mixed Effects Models
Qualitative parametersOrganic Mean ± SEConventional Mean ± SEExplanatory variablest-valued.f.P-value
Number of insect taxa7·80 ± 0·577·4 ± 0·65Farm typeNS
Number of plant species6·30 ± 0·674·5 ± 0·37Farm type3·3290·009
Number of interactions14·30 ± 1·6010·30 ± 1·10Farm type2·33590·044
   Flower abundanceNS
   Farm type × flower abundanceNS
Quantitative parameters
Generality2·07 ± 0·221·61 ± 0·13Farm type−2·15170·067
Flower abundanceNS
Farm type × flower abundance2·6070·035
Vulnerability3·03 ± 0·212·92 ± 0·32Farm type−2·56180·034
Insect abundance3·98580·004
Farm type × insect abundanceNS
Connectance0·19 ± 0·010·19 ± 0·01Farm typeNS
Specialization asymmetry−0·21 ± 0·05−0·20 ± 0·08Farm type2·41380·042
Insect abundance−3·49680·008
Flower abundanceNS
Insect niche overlap0·48 ± 0·040·42 ± 0·09Farm typeNS
image

Figure 2.  One of 10 pairs of organic and conventional insect–flower interaction networks. The bottom levels of each network show plant species (numbered 1–4) visited by insect taxa listed on the top levels. Plant bar thickness is in proportion to the frequency of visitation by insects. The lines between the two levels represent interactions between insect and plant species with line thickness indicating interaction frequency. Floral abundance is displayed in a composite column graph beneath each network as percentage of floral units of the top five most abundant species (representing 98% of total flower abundance at these two farms) in transects. Plant species are: 1 = Trifolium repens; 2 = Ranunculus repens L.; 3 = Rubus fruticosus agg.; 4 = Heracleum sphondylium L.; 5 = Bellis perennis. Note that full networks (containing bees/butterflies/hoverflies and all their observed interactions with flowers on a farm during the study period) are displayed. Any plant species in the composite column graphs but not in the networks were not observed receiving visits by bees, hoverflies or butterflies in this farm pair.

Download figure to PowerPoint

Organic and conventional networks were quite similar in terms of topology and structure. Quantitative connectance and niche overlap for plants and insects were the same in organic and conventional networks. Niche overlap was high for insects and low for plants. Although the majority of networks (65%) were found to have moderately asymmetrical structures (with plant species being more specialized than insect taxa), organic networks were significantly more asymmetrical than conventional networks which varied widely in structure (ranging from symmetrical to highly asymmetrical). Specialization asymmetry was significantly related to insect abundance on farms.

Pollination and farming system

Fruit set of C. monogyna flowers was significantly higher on organic compared with conventional farms for both open pollinated (= 0·002) and supplementally pollinated (< 0·001) flowers (Fig. 3). Crataegus monogyna flowers were found to be pollen limited (supplemental fruit set > open fruit set) on organic farms (= 0·006) but not on conventional farms (= 0·423) (Fig. 3).

image

Figure 3.  Mean (+SE) percentage of Crataegus monogyna flowers that set fruit after being either supplementally pollinated or left as open pollinated controls on organic and conventional farms. Letters above the bars indicate significant differences among treatments.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information

Insect–flower interaction networks and farming system

Insect–flower interaction networks on organic farms were bigger and differed in composition from networks on conventional farms. Insects visited more plant species (had higher generality) as a result of higher floral abundance on organic farms. In addition, plant species received more insect visits (higher vulnerability) on organic farms, which was positively influenced by insect abundance. Intensive grassland networks differed in size and composition, and particularly in floral abundance, depending on management (organic vs. conventional). To date, differences in food web size and structure (plant, herbivore, parasitoid) have been described for organic and conventional mixed arable farms only (Macfadyen et al. 2009).

All insect–flower interaction networks had low to moderately asymmetrical structures, with more specialized plant species than insects. This is in accordance with previous work, although our networks are less asymmetric than generally found elsewhere (e.g. Bascompte, Jordano & Olesen 2006; Vázquez et al. 2007). Interestingly, networks on organic farms were less variable in structure and more asymmetrical than those on conventional farms. More stable associations between mutualistic communities can be generated by interaction asymmetries (Bascompte, Jordano & Olesen 2006; Kaiser-Bunbury et al. 2010) because it is more beneficial for a specialist species to rely on a generalist species (asymmetry) rather than on another specialist (symmetry), in case the other specialist becomes less reliable and both specialist partners are then vulnerable to co-extinction (Blüthgen 2010). Therefore, the networks on organic farms in our study may be more stable than those on conventional farms. Differences in specialization asymmetries may be partly related to species richness and abundance, for example: floral abundance can influence interaction strength asymmetry (Vázquez et al. 2007), and we found more flowers and bees on organic farms.

However, we did not find a relationship between floral abundance and specialization asymmetry (similar to interaction strength asymmetry; Vázquez et al. 2007; Dormann et al. 2009; Blüthgen 2010), which suggests that floral abundance is not the most important determinant of insect–flower interaction network structure in grasslands. The quality of floral reward (i.e. its nutritional suitability for each consumer or ‘trait complementarity’) and its accessibility (to each consumer, i.e. the presence/absence of ‘exploitation barriers’), may be important as well as quantity (Stang et al. 2009). For example, Rubus fruticosus agg. flowers were not as abundant as other species, but attracted the highest proportion of insect visits, possibly because its pollen contains high percentages of essential amino acids and protein (Hanley et al. 2008), and large volumes of nectar are produced (Gyan & Woodell 1987b) in open flowers that are accessible to a variety of flower visitors (Muller 1881).

As insect (feeding and non-feeding) abundance increased, plants apparently became more specialized. This could be the result of some taxa foraging on previously less-favoured plants when visitor abundance was higher (Fontaine et al. 2008) because these plants became relatively rewarding in the presence of increased competition. However, not all insect taxa would be able to do this (because of trait complementarity and exploitation barriers; Stang et al. 2009), resulting in apparent increases in plant specialization as new plants are included in the network, but only by those insects that are able to exploit them.

The stability of insect–flower communities is thought to increase with a high degree of redundancy in insects (Memmott, Waser & Price 2004) because if a taxon is redundant in a network (has a similar interaction pattern to other species of the same trophic level) then its loss will not greatly destabilize the system. We found that there was a medium level of overlap (c. 45%), in both farm types, between the niches of insect taxa in our networks and the ratio of plants to insects was lower than that found in other networks (usually 1 : 4; Memmott, Waser & Price 2004; Jordano, Bascompte & Olesen 2006; Vázquez et al. 2009) which implies an inadequate level of insect redundancy (Memmott, Waser & Price 2004). All networks in our study were quite small, containing on average only 20 species. Even though our networks contain plant species that can reproduce without pollinators (e.g. R. fruticosus agg.), they will not be immune to insect taxa loss because inadequate insect redundancy and small network size may actually facilitate alterations in the system (i.e. changes in floral abundance, insect behaviour or abundance) resulting in increased vulnerability of insect-pollinated plants.

Quantitative connectance on organic and conventional farms was similar with networks dominated by well-connected generalist taxa, such as B. lucorum agg., Eristalis hoverflies and R. fruticosus agg. Removal of the most highly connected species (plant or insect) can cause rapid network collapse (Memmott, Waser & Price 2004) and is a likely event in intensive grasslands particularly in relation to plant species loss. Management activities such as silage and hedge cutting, topping (standardising grass height by cutting to an acceptable post-grazing height of c. 5 cm) and intensive grazing may remove the flowers of many plant species, effectively causing their ‘extinction’ from the network (at least temporarily). This effect may be exacerbated in our networks as the species pool is small, so the opportunity for insects to switch interaction partners (re-wiring) is low and therefore the networks could be greatly destabilized (Kaiser-Bunbury et al. 2010). Although organic farming can increase the size and specialization asymmetry of networks, the differences may be still too small to improve network stability and secure against future perturbations. Therefore, efforts should be made to retain and improve biodiversity by including measures in agri-environmental schemes that seek to maintain a constant floral abundance in intensive farms, particularly in relation to species that yield the highest reward for pollinators.

Flower visitors and farming system

As in previous studies (Rundlof, Nilsson & Smith 2008; Potts et al. 2009), the higher abundance and evenness of bees on organic farms was probably because of a higher abundance of flowers. The latter is likely to be a result of lower stocking densities, which alleviates grazing pressure and allows time for flowers to emerge. In addition, nitrogen-fixing legumes, such as Trifolium species are encouraged in organic systems as an alternative to chemical nitrogen fertilizers (Ledgard, Penno & Sprosen 1999) and were abundant in our organic farms. Trifolium species provide important food resources for bees (Pywell et al. 2005) as their pollen contains high percentages of essential amino acids and protein (Hanley et al. 2008). On conventional farms, B. perennis was the most abundant flower species, but it produces small quantities of low-sugar nectar (Schultz & Dlugosch 1999) and is not a valuable food source for bees (Fussell & Corbet 1992).

Hoverflies respond to habitat change differently to bees (Biesmeijer et al. 2006; Kleijn & van Langevelde 2006; Jauker et al. 2009). Hoverfly genera evenness (but not abundance or richness) was higher on our organic grasslands but, in contrast to previous studies (Kleijn & van Langevelde 2006; Meyer, Jauker & Steffan-Dewenter 2009) and to our findings for bees, there was no relationship between hoverflies and floral abundance, despite adults requiring nectar and pollen as food (Haslett 1989). Since some hoverfly species also feed on graminoid (grasses, sedges, rushes) pollen (Branquart & Hemptinne 2000; Speight 2008), and we did not count graminoid flowers, the lack of association in our study may be misleading. Unlike bees, hoverfly larvae do not require floral resources (Speight 2008). Therefore, the availability of larval microhabitats and food resources may be more important than floral abundance and requires further research.

Bees and hoverflies were significantly more associated with field edges than field centres, despite flower availability in the centres of fields. This may be because bees tend to follow linear features, possibly using them as navigational aids (Cranmer 2004). Wind speed and field boundary height are known to restrict hoverfly movement and cause them to accumulate near field boundaries (Wratten et al. 2003), for example, hoverfly evenness was negatively affected by wind speed in the present study. Hedgerows also offer the best food resources over the season (Jacobs et al. 2009) and give protection from predators and disturbance by livestock. Efforts to retain existing hedgerows and promote re-planting should be encouraged.

Pollination and farming system

Most pollination network studies have not addressed the links between observed insect assemblages and pollination (Ricketts et al. 2008). We used a single model species to gain a better insight into the effects of farm management on pollination. Clearly, pollination success would best be recorded for as many species in the community as possible, but owing to large variations in the presence of suitable plant species among sites, we could only use a single species. Despite this limitation, our study demonstrates farm management effects on pollination. Fruit set of openly pollinated C. monogyna flowers was significantly higher on organic farms. Here, we found more insect visitors, which presumably facilitated higher levels of pollination. Similar results have been found for canola crops (Brassica rapa and Brassica napus) (Morandin & Winston 2005). Crataegus monogyna was found to be pollen limited on organic farms, which is in accordance with work in the UK (Jacobs et al. 2009). This suggests that pollination levels, although higher than those on conventional farms, are still inadequate in organic grasslands.

In our study, fruit set of supplementally pollinated flowers was significantly lower, and was apparently not pollen limited, on conventional farms indicating that other factors are limiting. Nutrient limitation is possible, but no significant differences in pH, phosphorous, potassium and magnesium levels were found between organic and conventional field edges at the same study sites (E. F. Power, unpublished data). Alternatively, tree damage through cutting, spray drift from or direct application of herbicides (to control weeds) may reduce the fecundity of C. monogyna on conventional farms.

Summary of recommendations for management and policy

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information

Insect–flower interaction networks in organic grasslands were found to be larger than their conventional counterparts, but overall, the networks in our study (organic and conventional) were small and relatively unstable compared to networks in species-rich habitats. This is likely to be the result of disturbance from intensive management and is concerning because the maintenance of some level of pollination in intensively farmed landscapes is dependent on a network of very few species. However, organic farming was found to provide increased floral resources that attract more pollinating insects and pollination success was higher on farms under organic management. More research into the interacting effects of landscape and organic dairy farming is required (Gabriel et al. 2010), but our findings are an important starting point and have implications for how we perceive intensive grasslands. They are areas with some potential for pro-active conservation. The biodiversity that remains in intensive grasslands performs an essential role in the functioning of agricultural ecosystems and should be conserved and improved upon.

Organic dairy farming should be encouraged and supported particularly where intensive grasslands are widespread. Only 12 organic dairy farms in the Republic of Ireland met our selection criteria in 2008, 10 of which we surveyed, demonstrating the low uptake of organic dairy farming. Certain beneficial organic practices (for pollinators) can also be integrated into conventional farm management plans at little extra cost, for example, sowing Trifolium species to increase bee abundance. Strategic management of the flowers of important plant species for pollinators should be introduced into agri-environmental schemes, particularly in relation to hedgerow and pasture management. We confirm the value of hedgerows for pollinators and encourage hedgerow protection and restoration. This study also shows that more research is needed into hoverfly behaviour, dispersal and the influence of larval habitat and food requirements on adult distribution before realistic conservation recommendations can be made for this group. Ultimately, organic dairy farming, although not the solution in its present form, can improve prospects for pollinators in intensive grasslands.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information

We thank all the farmers for allowing us to sample in their fields; Doreen Gabriel, Daniel Kelly, Leanne Casey and Dara Stanley for advice and comments; David Rockett for help with fieldwork; and Christopher Kaiser-Bunbury, an anonymous reviewer and two journal editors for very helpful suggestions. This project, part of a collaborative project led by Michael O’Donovan, Teagasc, was funded by the Irish Department of Agriculture, Fisheries and Food Research Stimulus Fund (RSF), funded under the National Development Plan (2007–2013).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information
  • Bascompte, J., Jordano, P. & Olesen, J.M. (2006) Asymmetric coevolutionary networks facilitate biodiversity maintenance. Science, 312, 431433.
  • Bates, D. & Sarkar, D. (2006) Lme4: Linear Mixed-Effects Models Using S4 Classes. R package version 0.9975-10, http://cran.r-project.org/web/packages/lme4/index.html (accessed 6 January 2011).
  • Bengtsson, J., Ahnstrom, J. & Weibull, A.C. (2005) The effects of organic agriculture on biodiversity and abundance: a meta-analysis. Journal of Applied Ecology, 42, 261269.
  • Bersier, L.-F., Banasek-Richter, C. & Cattin, M.-F. (2002) Quantitative descriptors of food-web matrices. Ecology, 83, 23942407.
  • Biesmeijer, J.C., Roberts, S.P.M., Reemer, M., Ohlemuller, R., Edwards, M., Peeters, T., Schaffers, A.P., Potts, S.G., Kleukers, R., Thomas, C.D., Settele, J. & Kunin, W.E. (2006) Parallel declines in pollinators and insect-pollinated plants in Britain and the Netherlands. Science, 313, 351354.
  • Blüthgen, N. (2010) Why network analysis is often disconnected from community ecology: a critique and an ecologist’s guide. Basic and Applied Ecology, 11, 185195.
  • Blüthgen, N., Menzel, F., Hovestadt, T., Fiala, B. & Blüthgen, N. (2007) Specialization, constraints, and conflicting interests in mutualistic networks. Current Biology, 17, 341346.
  • Branquart, E. & Hemptinne, J.L. (2000) Selectivity in the exploitation of floral resources by hoverflies (Diptera: Syrphinae). Ecography, 23, 732742.
  • Cranmer, L. (2004) The influence of linear landscape features on pollinator behaviour. Unpublished PhD thesis, University of Northampton, Northampton.
  • Department of Agriculture (2009) Fact Sheet on Irish Agriculture. Department of Agriculture Fisheries and Food, Dublin, Ireland.
  • Dicks, L.V., Corbet, S.A. & Pywell, R.F. (2002) Compartmentalization in plant-insect flower visitor webs. Journal of Animal Ecology, 71, 3243.
  • Dormann, C.F., Fründ, J., Blüthgen, N. & Gruber, B. (2009) Indices, graphs and null models: analyzing bipartite ecological networks. The Open Ecology Journal, 2, 724.
  • Edwards, M. & Jenner, M. (2005) Field Guide to the Bumblebees of Great Britain and Ireland. Ocelli, London, UK.
  • Fontaine, C., Collin, C.L. & Dajoz, I. (2008) Generalist foraging of pollinators: diet expansion at high density. Journal of Ecology, 96, 10021010.
  • Free, J.B. (1993) Insect Pollination of Crops, 2nd edn. Academic Press, London, UK.
  • Fussell, M. & Corbet, S.A. (1992) Flower usage by bumble-bees: a basis for forage plant management. The Journal of Applied Ecology, 29, 451465.
  • Gabriel, D., Sait, S.M., Hodgson, J.A., Schmutz, U., Kunin, W.E. & Benton, T.G. (2010) Scale matters: the impact of organic farming on biodiversity at different spatial scales. Ecology Letters, 13, 858869.
  • Grime, J.P., Hodgson, J.G. & Hunt, R. (2007) Comparative Plant Ecology: A Functional Approach to Common British Species, 2nd edn. Castlepoint Press, London, UK.
  • Gyan, K.Y. & Woodell, S.R.J. (1987a) Analysis of Insect Pollen Loads and Pollination Efficiency of Some Common Insect Visitors of Four Species of Woody Rosaceae. Functional Ecology, 1, 269274.
  • Gyan, K.Y. & Woodell, S.R.J. (1987b) Nectar production, sugar content, amino acids and potassium in Prunus spinosa L., Crataegus monogyna Jacq. and Rubus fruticosus L. at Wytham, Oxfordshire. Functional Ecology, 1, 251259.
  • Hanley, M.E., Franco, M., Pichon, S., Darvill, B. & Goulson, D. (2008) Breeding system, pollinator choice and variation in pollen quality in British herbaceous plants. Functional Ecology, 22, 592598.
  • Haslett, J.R. (1989) Adult feeding by holometabolous insects: pollen and nectar as complementary nutrient sources for Rhingia campestris (Diptera: Syrphidae). Oecologia, 81, 361363.
  • Jacobs, J.H., Clark, S.J., Denholm, I., Goulson, D., Stoate, C. & Osborne, J.L. (2009) Pollination biology of fruit-bearing hedgerow plants and the role of flower-visiting insects in fruit-set. Annals of Botany, 104, 13971404.
  • Jauker, F., Diekotter, T., Schwarzbach, F. & Wolters, V. (2009) Pollinator dispersal in an agricultural matrix: opposing responses of wild bees and hoverflies to landscape structure and distance from main habitat. Landscape Ecology, 24, 547555.
  • Jordano, P., Bascompte, J. & Olesen, J.M. (2006) The ecological consequences of complex topology and nested structure in pollination webs. Plant-Pollinator Interactions: From Specialization to Generalization (eds N.M.Waser & J.Ollerton), pp. 173199. University of Chicago Press, Chicago, IL, USA.
  • Kaiser-Bunbury, C.N., Memmott, J. & Müller, C.B. (2009) Community structure of pollination webs of Mauritian heathland habitats. Perspectives in Plant Ecology, Evolution and Systematics, 11, 241254.
  • Kaiser-Bunbury, C.N., Muff, S., Memmott, J., Muller, C.B. & Caflisch, A. (2010) The robustness of pollination networks to the loss of species and interactions: a quantitative approach incorporating pollinator behaviour. Ecology Letters, 13, 442452.
  • Kleijn, D. & van Langevelde, F. (2006) Interacting effects of landscape context and habitat quality on flower visiting insects in agricultural landscapes. Basic and Applied Ecology, 7, 201214.
  • Knight, T.M., Steets, J.A., Vamosi, J.C., Mazer, S.J., Burd, M., Campbell, D.R., Dudash, M.R., Johnston, M.O., Mitchell, R.J. & Ashman, T.-L. (2005) Pollen limitation of plant reproduction: pattern and process. Annual Review of Ecology, Evolution, and Systematics, 36, 467497.
  • Ledgard, S.F., Penno, J.W. & Sprosen, M.S. (1999) Nitrogen inputs and losses from clover/grass pastures grazed by dairy cows, as affected by nitrogen fertilizer application. The Journal of Agricultural Science, 132, 215225.
  • Macfadyen, S., Gibson, R., Polaszek, A., Morris, R.J., Craze, P.G., Planque, R., Symondson, W.O.C. & Memmott, J. (2009) Do differences in food web structure between organic and conventional farms affect the ecosystem service of pest control? Ecology Letters, 12, 229238.
  • Memmott, J., Waser, N.M. & Price, M.V. (2004) Tolerance of pollination networks to species extinctions. Proceedings of the Royal Society B-Biological Sciences, 271, 26052611.
  • Meyer, B., Jauker, F. & Steffan-Dewenter, I. (2009) Contrasting resource-dependent responses of hoverfly richness and density to landscape structure. Basic and Applied Ecology, 10, 178186.
  • Morandin, L.A. & Winston, M.L. (2005) Wild bee abundance and seed production in conventional, organic, and genetically modified canola. Ecological Applications, 15, 871881.
  • Muller, H. (1881) Alpenblumen. Engelmann, Leipzig, Germany.
  • Murray, T.E., Fitzpatrick, U., Brown, M.J.F. & Paxton, R.J. (2008) Cryptic species diversity in a widespread bumble bee complex revealed using mitochondrial DNA RFLPs. Conservation Genetics, 9, 653666.
  • Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D. & the R Core team (2009) Nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1-96. http://CRAN.R-project.org/package=nlme (accessed 5 January 2011).
  • Plantureux, S., Peeters, A. & McCracken, D. (2005) Biodiversity in intensive grasslands: effect of management, improvement and challenges. Agronomy Research, 3, 152164.
  • Pollard, E. (1977) A method for assessing changes in abundance of butterflies. Biological Conservation, 12, 115134.
  • Potts, S.G., Woodcock, B.A., Roberts, S.P.M., Tscheulin, T., Pilgrim, E.S., Brown, V.K. & Tallowin, J.R. (2009) Enhancing pollinator biodiversity in intensive grasslands. Journal of Applied Ecology, 46, 369379.
  • Pywell, R.F., Warman, E.A., Carvell, C., Sparks, T.H., Dicks, L.V., Bennett, D., Wright, A., Critchley, C.N.R. & Sherwood, A. (2005) Providing foraging resources for bumblebees in intensively farmed landscapes. Biological Conservation, 121, 479494.
  • R Development Core Team (2007) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, available at: http://www.R-project.org (accessed 5 January 2011).
  • Ricketts, T.H., Regetz, J., Steffan-Dewenter, I., Cunningham, S.A., Kremen, C., Bogdanski, A., Gemmill-Herren, B., Greenleaf, S.S., Klein, A.M., Mayfield, M.M., Morandin, L.A., Ochieng, A. & Viana, B.F. (2008) Landscape effects on crop pollination services: are there general patterns? Ecology Letters, 11, 499515.
  • Rundlof, M., Nilsson, H. & Smith, H.G. (2008) Interacting effects of farming practice and landscape context on bumble bees. Biological Conservation, 141, 417426.
  • Schultz, C.B. & Dlugosch, K.M. (1999) Nectar and hostplant scarcity limit populations of an endangered Oregon butterfly. Oecologia, 119, 231238.
  • Shackleton, J., Nash, D. & Lewington, R. (2004) A Guide to the Butterflies of Ireland. The Dublin Naturalists’ Field Club, Dublin, Ireland.
  • Speight, M.C.D. (2008) Database of Irish Syrphidae (Diptera). Irish Wildlife Manuals, No. 36. National Parks and Wildlife Service, Department of Environment, Heritage and Local Government, Dublin, Ireland.
  • Stace, C.A. (2010) New Flora of the British Isles, 3rd edn. Cambridge University Press, Cambridge, UK.
  • Stang, M., Klinkhamer, P.G.L., Waser, N.M., Stang, I. & van der Meijden, E. (2009) Size-specific interaction patterns and size matching in a plant-pollinator interaction web. Annals of Botany, 103, 14591469.
  • Steffan-Dewenter, I., Potts, S.G. & Packer, L. (2005) Pollinator diversity and crop pollination services are at risk. Trends in Ecology & Evolution, 20, 651652.
  • Stubbs, A. (2002) British Hoverflies: An Illustrated Identification Guide. British Entomological and Natural History Society, Reading, UK.
  • Sugiura, N. (1996) Pollination of the orchid Epipactis thunbergii by syrphid flies (Diptera: Syrphidae). Ecological Research, 11, 249255.
  • Tylianakis, J.M., Tscharntke, T. & Lewis, O.T. (2007) Habitat modification alters the structure of tropical host-parasitoid food webs. Nature, 445, 202205.
  • Vance, N.C., Bernhardt, P. & Edens, R.M. (2004) Pollination and seed production in Xerophyllum tenax (Melanthiaceae) in the Cascade Range of central Oregon. American Journal of Botany, 91, 20602068.
  • Vázquez, D.P., Melián, C.J., Williams, N.M., Blüthgen, N., Krasnov, B.R. & Poulin, R. (2007) Species abundance and asymmetric interaction strength in ecological networks. Oikos, 116, 11201127.
  • Vázquez, D.P., Blüthgen, N., Cagnolo, L. & Chacoff, N.P. (2009) Uniting pattern and process in plant-animal mutualistic networks: a review. Annals of Botany, 103, 14451457.
  • Vickery, J.A., Tallowin, J.R., Feber, R.E., Asteraki, E.J., Atkinson, P.W., Fuller, R.J. & Brown, V.K. (2001) The management of lowland neutral grasslands in Britain: effects of agricultural practices on birds and their food resources. Journal of Applied Ecology, 38, 647664.
  • Westphal, C., Bommarco, R., Carré, G., Lamborn, E., Morison, N., Petanidou, T., Potts, S.G., Roberts, S.P.M., Szentgyörgyi, H., Tscheulin, T., Woyciechowski, M., Biesmeijer, J.C., Kunin, W.E., Settele, J. & Steffan-Dewenter, I. (2008) Measuring bee diversity in different European habitats and biogeographical regions. Ecological Monographs, 78, 653671.
  • Westrich, P. (1989) Die Wildbienen Baden-Württembergs. Eugen Ulmer, Stuttgart, Germany.
  • Wolf, S., Rohde, M. & Moritz, R.F.A. (2010) The reliability of morphological traits in the differentiation of Bombus terrestris and B. lucorum (Hymenoptera: Apidae). Apidologie, 41, 4553.
  • Wratten, S.D., Bowie, M.H., Hickman, J.M., Alison, M.E., Sedcole, J.R. & Jason, M.T. (2003) Field boundaries as barriers to movement of hover flies (Diptera: Syrphidae) in Cultivated Land. Oecologia, 134, 605611.
  • Zuur, A.F., Ieno, E.N., Walker, N.J., Saveliev, A.A. & Smith, G.M. (2009) Mixed Effects Models and Extensions in Ecology With R. Springer, New York, NY, USA.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Summary of recommendations for management and policy
  8. Acknowledgements
  9. References
  10. Supporting Information

Figure S1. Distribution of the ten farm pairs within Ireland.

Figure S2. Floral units observed in organic and conventional field edge and centre transects.

Figures S3–S11. Organic and conventional insect–flower interaction webs – farm pair 2 to 10.

Table S1. Summary of management activities in organic and conventional dairy farms

Table S2. Significant variables characterizing bee and hoverfly abundance, richness and evenness in the edges and centres of organic and conventional fields.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JPE_1949_sm_FigS1.pdf48KSupporting info item
JPE_1949_sm_FigS2.pdf37KSupporting info item
JPE_1949_sm_FigS3-11.pdf134KSupporting info item
JPE_1949_sm_TableS1.pdf14KSupporting info item
JPE_1949_sm_TableS2.pdf13KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.