- Email: [email protected]
Impact of wildflower strips on biological control of cabbage lepidopterans
Agriculture, Ecosystems and Environment 129 (2009) 310–314
Contents lists available at ScienceDirect
Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee
Impact of wildflower strips on biological control of cabbage lepidopterans L. Pfiffner a,*, H. Luka a,c, C. Schlatter a, A. Juen b, M. Traugott b a
Research Institute of Organic Agriculture, Ackerstraße, 5070 Frick, Switzerland University of Innsbruck, Institute of Ecology, Mountain Agriculture Research Unit, Technikerstrasse 25, 6020 Innsbruck, Austria c University of Basel, Department of Environmental Sciences, Institute of Biogeography, St. Johanns-Vorstadt 10, CH-4056 Basel, Switzerland b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 June 2008 Received in revised form 3 October 2008 Accepted 6 October 2008 Available online 14 November 2008
In a 2-year experiment we investigated whether wildflower strips can be used to enhance the control of cabbage moth, Mamestra brassicae L., and cabbage white butterfly, Pieris rapae L. At two sites, including six organically cultivated fields, M. brassicae egg parasitism and predation rates were determined along with an assessment of larval parasitism rates in M. brassicae and P. rapae using a DNA-based approach. Within each field, plots with and without wildflower strips were sampled and a grid design of 3 m 3 m was used to analyze the spatial pattern of parasitism. The provision of wildflower strips provided an idiosyncratic effect on the control of lepidopterans: parasitism rates in M. brassicae eggs and larvae were not affected, whereas parasitism rates of larval P. rapae were significantly enhanced by the wildflower strips at one of the two sites. Moreover, at one site predation rates on M. brassicae eggs were significantly enhanced in the wildflower strip plots. Geostatistical analysis showed no distinct spatial patterns in parasitism rates. These results demonstrate that the provision of wildflower strips does not necessarily enhance biological control of lepidopteran cabbage pests and suggest that site-specific environmental factors strongly affect the impact of wildflower strips. ß 2008 Elsevier B.V. All rights reserved.
Keywords: Conservation biological control Parasitoids Diagnostic PCR Habitat manipulation Mamestra brassicae Pieris rapae
1. Introduction In agricultural landscapes dominated by large arable monocultures many beneficial arthropod species tend to suffer from a lack of non-pest food sources such as nectar and pollen, accompanied by a lack of sites for shelter, hibernation, mating and nesting (Jervis and Heimpel, 2005; Gurr et al., 2004). Vegetative buffers and uncropped field margins provide many of these vital resources, helping to sustain beneficial arthropods in agricultural land (Landis et al., 2000; Pfiffner and Luka, 2000). Wildflower strips can provide means to enhance natural enemy populations in arable land and to increase their regulative power on pests (Lavandero et al., 2005; Winkler et al., 2006). Within the Swiss agro-environmental program, species-rich wildflower strips are recommended to support wildlife conservation in arable land (Pfiffner and Wyss, 2004; Herzog and Walter, 2005). The wildflower seed mixtures include annual, biennial, and perennial plants; all of them are native to Switzerland to avoid detrimental effects on the indigenous flora. Cabbage crops are attacked by a variety of insects including cabbage moth, Mamestra brassicae L., and cabbage white, Pieris
rapae L. Their caterpillars feed on cruciferous and related food plants, defoliating and contaminating the plants with large quantities of faecal matter, thus being a serious pest of cultivated Cruciferae (Hill, 1987). M. brassicae and P. rapae are both attacked by a broad spectrum of natural enemies: eggs are parasitized by Trichogramma sp. and Telenomus sp., the larvae of P. rapae are attacked by the gregarious endoparasitoid Cotesia glomerata (L.) and the solitary wasp Cotesia rubecula (Marshall) (both Hymenoptera: Braconidae), whereas the main parasitoid of larval M. brassicae is the solitary endoparasitoid Microplitis mediator (Haliday) (Hymenoptera: Braconidae) (Harvey et al., 1999; Lauro et al., 2005). Moreover, eggs and larvae are preyed on by generalist predators such as spiders, chrysopids, staphylinids and carabids (Johansen, 1997). The aims of this 2-year study conducted at two farms in Switzerland were three-fold: (i) to assess whether predation rates on M. brassicae eggs can be enhanced by sown wildflower strips, (ii) to test whether the provision of wildflower strips increases parasitism rates in M. brassicae and P. rapae and (iii) to assess whether wildflower strips influence the spatial patterns in parasitism. 2. Material and methods
* Corresponding author. Tel.: +41 62 8657246; fax: +41 62 8657273. E-mail address: lukas.pfiffner@fibl.org (L. Pfiffner). 0167-8809/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.agee.2008.10.003
The experiments were conducted in north-western Switzerland at one organic farm in Bibern, canton Solothurn, and another one at
L. Pfiffner et al. / Agriculture, Ecosystems and Environment 129 (2009) 310–314
Murimoos, canton Bern. To assess the effects of sown wildflower strips on pest control, we monitored egg and larval parasitism as well as egg-predation rates in six cabbage fields. Each field was divided into two blocks, each block with a cabbage plot with and without sown wildflower strips. The plot size was 45 m 25 m. In 2004, two fields per site which bordered to the opposite side of the same wildflower strip were sampled, one field per site in 2005. The field margins of the control plots were covered by plastic sheets in order to prevent weeds from establishing. At Murimoos cabbage (Brassica oleracea var. capitata) was planted on 26 May at one side and on 8 June 2004 at the other side of the wildflower strip. In 2005 cabbage was planted at one side only on 20 June. At Bibern broccoli (Brassica oleracea var. silvestris) was planted on 3 May 2004 on one side of the wildflower strip, whereas cabbage was planted on the other side on 17 May. In 2005 cabbage was planted at one side only on 23 May. A 3 m 3 m grid design was used to analyse the spatial pattern in parasitism; 36 or 48 labelled plants per block were used for assessing egg parasitism (up to 18 m from the strip) and larval parasitism (up to 24 m from the strip), respectively. The mixture of the wildflower strip used in this study is officially recommended within the Swiss agro-environmental program. Wildflower strips were sown with a mixture of 24 wild flower species in spring or autumn preceding the investigations. The size of strips were 3 m 35 m. During the sampling periods six to 14 plant species were blooming (Table 1). Arable weeds occurred within the study fields at low level only, weed coverage ranging between 0.5% and 3%. No pesticides and mechanical control measures were applied during the experiments. Within each year cabbage was planted in new fields, with field size varying from 0.66 to 0.80 ha. 2.1. Assessing egg parasitism and predation To investigate parasitism and predation rates on eggs of M. brassicae, paper cards with egg batches originating from laboratory colonies were placed out in the fields. One egg batch each (average batch size 21 eggs) was pinned with a needle to the ground underneath a crop leaf at each of the 36 grid sampling points. Eggs were not older than 1 day and were placed out between 7–23 July 2004 and between 5–28 July 2005 at four dates each year. After a 3day-exposure period, all cards were collected and incubated in the laboratory for 4 weeks at 22 8C to rear egg parasitoids. Parasitism rates were calculated as the proportion of the number of parasitoids reared in relation to the number of eggs exposed. Egg predation rates were calculated by comparing the number of eggs per batch before and after field exposure; this was done in 2005 only. Using a binocular microscope, the number of eggs was counted and empty, damaged or missing eggs were considered as destroyed or removed by invertebrate predators. 2.2. Determining larval parasitism To determine the rates of hymenopteran parasitism in P. rapae and M. brassicae larvae a DNA-based approach was employed. Larvae of both lepidopteran species were collected at three dates between 20 July and 3 August 2005 (site Bibern: two dates only). At each date, all larvae found on 96 cabbage plants per treatment were collected; new plants were selected for each sampling date. Each larva was identified to species level, put in a 1.5 ml reaction tube and frozen at 28 8C for subsequent molecular analysis. The total DNA of each field-collected larva was extracted using the CTAB-protocol described in Juen and Traugott (2005). Small larvae were extracted as a whole, medium and large-sized larvae were cut in two or three pieces, which were subsequently
311
processed as separate samples. Within each batch of 30 samples, one negative extraction control was included to check for DNA carry-over contamination. Primer pairs S29/A27 and S31/A28 (Traugott et al., 2006) targeting C. glomerata and C. rubecula, respectively, were used in multiplex PCR for testing P. rapae larvae in polymerase chain reaction (PCR) for these two parasitoids (for reaction mix and cycling conditions see Traugott et al., 2006). To exclude false negative PCR results due to PCR inhibitors or DNA extraction failures, all DNA extracts of P. rapae which did not test positive for parasitoid DNA were tested with general primers (Folmer et al., 1994) to ensure their amplifyability. Each 20 ml PCR mix contained 200 mM dNTPs, 1 mM of each primer, 2 ml of 10x buffer, 3 mM MgCl2, 0.75 U Taq DNA polymerase (BioTherm, GeneCraft), 1 ml bovine serum albumin (20 mg/ml), and 2 ml of extracted DNA. The cycling conditions were 2 min at 94 8C, 30 cycles of 15 s at 94 8C, 15 s at 52 8C, 45 s at 72 8C and final elongation for 2 min at 72 8C. Based on these tests, eight larvae of P. rapae (n = 662) had to be excluded from the analysis. Mitochondrial cytochrome oxide subunit I sequences (GenBank accession numbers see Traugott et al., 2006) were used to design primer pairs S100/A100 (S100 (50 -30 ) TTTAGAATTAGGAATATCTGGGAA; A100 (50 –30 ) TAAACCTCTGTGTCCCAAAATT) and S121/A124 (S121 (50 –30 ) GAATTAGGAAACCCTGGATC; A124 (50 –30 ) AAAGAGTTAAAGAAGGGGGA) to amplify a 310 bp fragment from M. mediator and a 217 bp fragment from M. brassicae, respectively. The latter primer pair served as an internal control for amplification-failure in the multiplex PCR and lead to the exclusion of eight M. brassicae larvae (n = 153) from the analysis. Each 10 ml multiplex PCR mix contained 1.5 ml of extracted DNA, 1 multiplex PCR master mix (QIAGEN), 0.2 mM of each primer and 2.5 ml PCR-water. The cycling conditions were 15 min at 95 8C, 30 cycles of 30 s at 94 8C, 90 s at 60 8C, 60 s at 72 8C and final elongation for 10 min at 72 8C. The specificity of the diagnostic PCR assays was tested using the target parasitoid species, its host species the lepidopteran species Pieris brassicae L. and Plutella xylostella L. as well as Diadegma semiclausum Helle´n (Hymenoptera: Ichneumonidae; main parasitoid of P. xylostella), all being abundant species in cabbage fields. The PCR assays proved to be specific for the target species. All PCR assays included DNA of the target species as a positive control, DNA of the host as positive/negative control and PCR water allowing to detect DNA carry-over contamination. PCR products were loaded onto ethidium bromide-stained agarose gels, separated by size using electrophoresis and visualized on a UVtransiluminator. 2.3. Statistical analysis Mann–Whitney U-test was used for non-parametric data to test for differences between treatments in egg and larval parasitism rates on M. brassicae and P. rapae, and predation rates on M. brassicae eggs and pest densities using the pooled data of all sampling dates with the software SPSS version 10. Analyses were done separately with arcsin-transformed data for each field. Two blocks per field were used to test treatments wildflower vs. control. The analysis of the distance gradients from the wildflower strip into the field was performed using the pooled data of exposed eggs within the grid. Additionally, geostatistical analysis was applied as exploratory data analysis to supplement the statistical analysis described above and to generate hypotheses to be tested with non-spatial statistics. The inverse distance weighting (IDW) method was employed, providing a more robust interpolation algorithm than kriging and indicator kriging (Clark and Harper, 2000).
312
L. Pfiffner et al. / Agriculture, Ecosystems and Environment 129 (2009) 310–314
Fig. 3. Mean parasitism rates (S.E.M.) in caterpillars of Pieris rapae. Comparison of wildflower strip (+) and control () treatment at the two sites Bibern (B) and Murimoos (M) using Mann–Whitney U-test, *: significant difference, p < 0.05, ns: no significant difference, p > 0.05.
Fig. 1. Mean parasitism rates (S.E.M.) in Mamestra brassicae eggs in relation to the distance to the wildflower strip. Plots with (*) and without (*) wildflower strips at site Bibern (a) and Murimoos (b). Mann–Whitney U-test with significant difference (*) with p = 0.042.
3. Results In the 2004 experiments, placed-out sentinel eggs (n > 21,000) were parasitized by T. evanescens Westwood and Telenomus sp., the latter being significantly more abundant than T. evanescens. Overall egg parasitism rates did not differ between wildflower and control treatments nor between most distances to the wildflower strip (Fig. 1). At both sites, no influence of the treatment on the spatial pattern in egg parasitism could be found using geostatistical analysis. Only a general edge effect at Murimoos, with higher parasitism rates occurring along the field margin could be detected. In 2005, egg parasitism was very low (<1%) most likely due to low temperatures and rainy weather (data not shown), precluding any further analysis at both sites. Overall predation rates on M. brassicae eggs varied from 2.2% to 6.4%, with higher predation rates occurring close to field margins. At Bibern predation on M. brassicae eggs was significantly higher in plots adjacent to wildflower strips than in those bordering to control plots (Fig. 2), however, this pattern did not occur at Murimoos. Only lacewing larvae were observed to feed on egg batches in the field.
Fig. 2. Mean predation rates (S.E.M.) on 3-day-exposed eggs of Mamestra brassicae. Comparison of predation rates between wildflower strip (+) and control () treatment at the two sites Bibern (B) and Murimoos (M) using Mann–Whitney U-test, *: significant difference, p < 0.05, ns: no significant difference, p > 0.05.
Larval densities of both M. brassicae and P. rapae did not significantly differ between treatments (Mann–Whitney U-test with p > 0.05). Mean caterpillar densities per plant (S.E.M.) varied in M. brassicae from 1.1 0.16 (Bibern) to 1.6 0.07 (Murimoos) and in P. rapae from 1.4 0.09 (Bibern) to 3.2 0.15 (Murimoos). High larval parasitism rates between 45% and 75% were detected in M. brassicae at both sites and in P. rapae at Murimoos. At Bibern, however, only one out of 106 collected P. rapae larvae was positive for parasitoid DNA (Fig. 3). At Murimoos, C. rubecula was the main parasitoid of P. rapae (six out of 548 caterpillars (1.09%) with another four yielding DNA of both Cotesia species). At Murimoos P. rapae larvae showed significantly higher parasitism rates in plots bordering wildflower strips than in control plots, whereas no such effect could be found at Bibern (Fig. 3). Parasitism rates in M. brassicae larvae did not differ significantly between treatments at both sites. 4. Discussion The provision of wildflower strips in organic cabbage cultures provided an idiosyncratic effect on the control of lepidopteran cabbage pests: parasitism rates of M. brassicae eggs and larvae as well as of P. rapae larvae at Bibern were not affected, whereas parasitism rates of larval P. rapae at Murimoos and predation rates on M. brassicae eggs were significantly enhanced by the wildflower strips. Only a few field studies have assessed the impact of flower strips on pest control in vegetable crops and their results are contradictory. Lavandero et al. (2005), Lee and Heimpel (2005), Winkler (2005) and Winkler et al. (2006) have shown that flower strips can increase parasitism rates in neighbouring fields, but there are also studies reporting on no differences in parasitism rates between fields with and without flower strips (e.g. Berndt et al., 2002). The ability of parasitoids to utilize floral resource subsidies is dependent on a wide range of factors, including flower morphology, colour, odour or the timing of nectar production (Heimpel and Jerwis, 2005). Considering all these aspects, only a few of the plant species used in the seed mixture such as C. cyanus, F. esculentum and D. carota may have benefitted the target parasitoids (Lee and Heimpel, 2005; Winkler, 2005). Although parasitism rates in the artificially exposed M. brassicae eggs ranged from 14% to 52%, the wildflower strips did not enhance egg parasitism rates. Moreover, parasitism rates did not change significantly with increasing distance to the field border in both plots with and without wildflower strips. Egg parasitoids cover only short distances by active dispersal (Fournier and Boivin, 2000)
L. Pfiffner et al. / Agriculture, Ecosystems and Environment 129 (2009) 310–314 Table 1 Seed mixture of wildflower strip and occurrence of blooming wild flowers during the sampling period in 2004 and 2005 at both sites. Sown species
Blooming plant species g/ha
Achillea millefolium L. Agrostemma githago L. Anthemis tinctoria L. Centaurea cyanus L. Centaurea jacea L. Cichorium intybus L. Daucus carota L. Dipsacus silvester L. Echium vulgare L. Fagopyrum esculentum Miller Hypericum perforatum L. Leucanthemum vulgare LAM. Legousia speculum-veneris L. Malva moschata L. Malva sylvestris L. Melilotus albus MED. Onobrychis viciifolia Scop. Origanum vulgare L. Papaver rhoeas L. Pastinaca sativa L. Silene alba L. Tanacetum vulgare L. Verbascum densiflorum L. Verbascum lychnitis L.
20 600 20 500 200 120 150 2 200 7245
Unsown dominant species Matricaria chamomilla L. Total number
Bibern 2004
Murimoos
x x x x
x x x
x
Bibern 2005
Murimoos
x x x x x x x r x r
r x x x
60 80
x
r x
30
r
r
20 60 20 600 60 150 80 100 3 50 30
x
r
x
r x
r
x
r
x
x
x
6
14
10
x 6
parasitism rates in M. brassicae eggs correlated positively with the area of woody habitats and pasture at the field level. Besides attracting predators and parasitoids, introducing nectar sources to a cropping system could also increase the density of pest insects which might benefit from the same food sources (Baggen et al., 1999; Zhao et al., 1992). No such effect was found in the present study: larval densities of M. brassicae and P. rapae were similar in plots with and without wild flower strips. In conclusion, our results indicate that the floral treatment using a multiple species mixture did not consistently improve the control of P. rapae and M. brassicae compared with a nearby control treatment under the small-scaled vegetable crop production situation in Switzerland. A more effective conservation biological control may be achieved by the provision of selective floral resources (Lavandero et al., 2006) and the creation of permanent non-crop habitats as e.g. woodlots and hedgerow (Den Belder et al., 2006). Acknowledgements
x
0
313
r: rare plant sp.; x > 1% cover.
but are able to recolonise from overwintering habitats such as pastures (Schneider et al., 2003). The predation on M. brassicae eggs was found to be higher on plants near to the wildflower strip. Wildflower strips are known to harbour many epigeic polyphagous arthropods during winter (Pfiffner and Luka, 2000), which can immigrate during spring into the adjacent arable crops (Holland et al., 2003), causing the high predation rates observed. Parasitism rates in larval M. brassicae were not affected by the wildflower strip, whereas in P. rapae larvae significantly higher parasitism rates were observed at one of the two sites. The latter finding is in accordance with the study of Lee and Heimpel (2005), finding that buckwheat (F. esculentum) strips increased parasitism rates in caterpillars of P. rapae. Both C. glomerata and C. rubecula are known to parasitize P. rapae; however, C. rubecula regularly outcompetes C. glomerata in this host (Geervliet et al., 2000). This was confirmed by our study as 0.7% of the caterpillars were parasitized by both Cotesia species; just 0.4% of them were parasitized by C. glomerata only. In a previous study using the same PCR assay (Traugott et al., 2006), no multiparasitism was detected in P. rapae and P. brassicae larvae collected at Murimoos in 2004, yet a total of only 157 larvae was screened. Interestingly, parasitism rates in P. rapae larvae were quite similar in 2004 (53.4%, n = 30) and in 2005 (53.4%, n = 548) at this site, suggesting that site-specific effects on larval parasitism occur. Permanent non-crop habitats may play a crucial role for parasitoids to build up their populations (Thies and Tscharnke, 1999). For example, the presence of woodlots and hedgerows can negatively affect the number of P. rapae larvae in arable land (Den Belder et al., 2006). Bianchi et al. (2005) found that predation and
This study was financially supported by Federal Office for the Environment, Federal Office for Agriculture, Fonds Landschaft Schweiz, Canton of Fribourg and Bern and MAVA Foundation, Sophie-Carl Binding Foundation. We thank H. Baur, K. van Achterberg K. Horstmann and J. C. Monje for the support in parasitoid determination. The authors are grateful to the farmers for providing the fields as well as for the comments on the manuscript provided by anonymous reviewers. References Baggen, L.R., Gurr, G.M., Meats, A., 1999. Flowers in tri-trophic systems: Mechanisms allowing selective exploitation by insect natural enemies for conservation biological control. Entomol. Exp. Appl. 91, 155–161. Berndt, L.A., Wratten, S.D., Hassan, P.G., 2002. Effects of buckwheat flowers on leafroller (Lepidoptera: Tortricidae) parasitoids in New Zealand vineyard. Agric. Forest Entomol. 4, 39–45. Bianchi, F.J.J.A., van Wingerden, W.K., Griffioen, A.J., van der Veen, M., van der Straten, M.J., Wegman, R.M.A., Meeuwsen, H.A.M., 2005. Landscape factors affecting the control of Mamestra brassicae by natural enemies in Brussels sprout. Agric. Ecosyst. Environ. 107, 145–150. Clark, I., Harper, W.V., 2000. Practical Geostatistics. Ecosse North America Ltd., Columbus, OH. Den Belder, J., Elderson, J., Schelling, G., 2006. Landscape effect on the abundance of lepidopteran pests in Brussel sprouts. IOBC/wprs Bulletin 29, 25–28. Folmer, O., Black, M., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Marine Biol. Biotech. 3, 294–299. Fournier, F., Boivin, G., 2000. Comparative dispersal of Trichogramma evanescens and Trichogramma pretiosum (Hymenoptera: Trichogrammatidae) in relation to environmental conditions. Popul. Ecol. 29, 55–63. Geervliet, J.B.F., Verdel, M.S.W., Snellen, H., Schaub, J., Dicke, M., Vet, L.E.M., 2000. Coexistence and niche segregation by Weld populations of the parasitoids Cotesia glomerata and C. rubecula in the Netherlands: predicting Weld performance from laboratory data. Oecologia 124, 55–63. Gurr, G.M., Wratten, S.D., Altieri, M.E., 2004. Ecological Engineering for Pest Management: Advances in Habitat Manipulation for Arthropods. CSIRO Publishing, Collingwood, Vic, Australia. Harvey, J.A., Jervis, M.A., Gols, R., Jiang, N., Lem, M.V., 1999. Development of the parasitoid Cotesia rubecula (Hymeroptera: Braconidae) in Pieris rapae and Pieris brassicae (Lepidoptera: Pieridae): evidence for host regulation. J. Ins. Physiol. 45, 173–182. Heimpel, G.E., Jerwis, M.A., 2005. Does nectar improve biocontrol by parasitoids? In: Wa¨ckers, F., van Rijn, P.C.,Bruin, J. (Eds.), Plant provided food for carnivorous insects: a protective Mutalism and its Applications. Chapman & Hall, London, UK, pp. 267–304.
Plant provided food for carnivorous insects ¨ komassnahmen, Bereich BiodiHerzog, F., Walter, T. (Eds.), 2005. Evaluation der O versita¨t, 56. Schriftenreihe FAL, pp 205. Hill, D.S., 1987. Agricultural Insect Pests of Temperate Regions and their Control. Cambridge Univ. Press, Cambridge.
314
L. Pfiffner et al. / Agriculture, Ecosystems and Environment 129 (2009) 310–314
Holland, J., Birkett, T., Begbie, M., Southway, S., Thomas, C.F.G., 2003. The spatial dynamics of predatory arhropods and the importance of crop and adjacent margin habitats. IOBC/wprs Bull. 26, 65–70. Jervis, M.A., Heimpel, G.E., 2005. Phytophagy. In: Jervis, M.A. (Ed.), Insects as natural enemies. A practical perspective. Springer, Dordrecht, pp. 525–555. Johansen, N.S., 1997. Mortality of eggs, larvae and pupae and larval dispersal of the cabbage moth, Mamestra brassicae, in white cabbage in south-eastern Norway. Entomol. Exp. Appl. 83, 347–360. Juen, A., Traugott, M., 2005. Detecting predation and scavenging by DNA gutcontent analysis: a case study using a soil insect predator–prey system. Oecologia 142, 344–352. Landis, D.A., Wratten, S.D., Gurr, G.M., 2000. Habitat management to conserve natural enemies of arthropod pests in agriculture. Ann. Rev. Entomol. 45, 175–201. Lauro, N., Kuhlmann, U., Mason, P.G., Holliday, N.J., 2005. Interaction of a solitary larval endoparasitoid, Microplitis mediator, with its host. Mamestra brassicae: host acceptance and host suitability. J. Appl. Entomol. 129, 567–573. Lavandero, B., Wratten, S., Didham, R., Gurr, G.M., 2006. Increasing floral diversity for selective enhancement of biological control agents: a double-edged sward? Basic Appl. Ecol. 7, 236–243. Lavandero, B., Wratten, S., Shishehbor, P., Worner, S., 2005. Enhancing the effectiveness of the parasitoid Diadegma semiclausum (Hellen): movement after use of nectar in the field. Biol. Control 34, 152–158. Lee, J.C., Heimpel, G.E., 2005. Impact of flowering buckwheat on lepidopteran pests and their parasitoids at two spatial scales. Biol. Control 34, 290–301.
Pfiffner, L., Luka, H., 2000. Overwintering of arthropods in soils of arable fields and adjacent semi-natural habitats. Agric. Ecosyst. Environ. 78, 215–222. Pfiffner, L., Wyss, E., 2004. Use of sown wildflower strips to enhance natural enemies of agricultural pests. In: Gurr, G., Wratten, S., Altieri, M. (Eds.), Ecological Engineering for Pest Management: Advances in Habitat Manipulation for Arthropods. CSIRO Publishing, Collingwood, Vic, Australia, pp. 165– 186. Schneider, C., Dover, J., Fry, G.L.A., 2003. Movement of two grassland butterflies in the same habitat network: the role of adult resources and size of the study area. Ecol. Entomol. 28, 219–227. Thies, C., Tscharnke, T., 1999. Landscape structure and biological control in agroecosystems. Science 6, 893–895. Traugott, M., Zangerl, P., Juen, A., Schallhart, N., Pfiffner, L., 2006. Detecting key parasitoids of lepidopteran pests by multiplex PCR. Biol. Control 39, 39–46. Winkler, K., 2005. Assessing the risks and benefits of flowering field edges. Strategic use of nectar sources to boost biological control. Thesis, Wageningen University, p. 118. Winkler, K., Wa¨ckers, F., Bukovinszkine-Kiss, G., van Lenteren, J.C., 2006. Sugar resources are vital for Diadegma semiclausum fecundity under field conditions. Basic Appl. Ecol. 7, 133–144. Zhao, J.Z., Ayers, G.S., Grafius, E.J., Stehr, F.W., 1992. Effects of neighboring nectarproducing plants on populations of pest lepidoptera and their parasitoids in broccoli plantings. The Great Lakes Entomologist 25, 253–258.
Contents lists available at ScienceDirect
Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee
Impact of wildflower strips on biological control of cabbage lepidopterans L. Pfiffner a,*, H. Luka a,c, C. Schlatter a, A. Juen b, M. Traugott b a
Research Institute of Organic Agriculture, Ackerstraße, 5070 Frick, Switzerland University of Innsbruck, Institute of Ecology, Mountain Agriculture Research Unit, Technikerstrasse 25, 6020 Innsbruck, Austria c University of Basel, Department of Environmental Sciences, Institute of Biogeography, St. Johanns-Vorstadt 10, CH-4056 Basel, Switzerland b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 June 2008 Received in revised form 3 October 2008 Accepted 6 October 2008 Available online 14 November 2008
In a 2-year experiment we investigated whether wildflower strips can be used to enhance the control of cabbage moth, Mamestra brassicae L., and cabbage white butterfly, Pieris rapae L. At two sites, including six organically cultivated fields, M. brassicae egg parasitism and predation rates were determined along with an assessment of larval parasitism rates in M. brassicae and P. rapae using a DNA-based approach. Within each field, plots with and without wildflower strips were sampled and a grid design of 3 m 3 m was used to analyze the spatial pattern of parasitism. The provision of wildflower strips provided an idiosyncratic effect on the control of lepidopterans: parasitism rates in M. brassicae eggs and larvae were not affected, whereas parasitism rates of larval P. rapae were significantly enhanced by the wildflower strips at one of the two sites. Moreover, at one site predation rates on M. brassicae eggs were significantly enhanced in the wildflower strip plots. Geostatistical analysis showed no distinct spatial patterns in parasitism rates. These results demonstrate that the provision of wildflower strips does not necessarily enhance biological control of lepidopteran cabbage pests and suggest that site-specific environmental factors strongly affect the impact of wildflower strips. ß 2008 Elsevier B.V. All rights reserved.
Keywords: Conservation biological control Parasitoids Diagnostic PCR Habitat manipulation Mamestra brassicae Pieris rapae
1. Introduction In agricultural landscapes dominated by large arable monocultures many beneficial arthropod species tend to suffer from a lack of non-pest food sources such as nectar and pollen, accompanied by a lack of sites for shelter, hibernation, mating and nesting (Jervis and Heimpel, 2005; Gurr et al., 2004). Vegetative buffers and uncropped field margins provide many of these vital resources, helping to sustain beneficial arthropods in agricultural land (Landis et al., 2000; Pfiffner and Luka, 2000). Wildflower strips can provide means to enhance natural enemy populations in arable land and to increase their regulative power on pests (Lavandero et al., 2005; Winkler et al., 2006). Within the Swiss agro-environmental program, species-rich wildflower strips are recommended to support wildlife conservation in arable land (Pfiffner and Wyss, 2004; Herzog and Walter, 2005). The wildflower seed mixtures include annual, biennial, and perennial plants; all of them are native to Switzerland to avoid detrimental effects on the indigenous flora. Cabbage crops are attacked by a variety of insects including cabbage moth, Mamestra brassicae L., and cabbage white, Pieris
rapae L. Their caterpillars feed on cruciferous and related food plants, defoliating and contaminating the plants with large quantities of faecal matter, thus being a serious pest of cultivated Cruciferae (Hill, 1987). M. brassicae and P. rapae are both attacked by a broad spectrum of natural enemies: eggs are parasitized by Trichogramma sp. and Telenomus sp., the larvae of P. rapae are attacked by the gregarious endoparasitoid Cotesia glomerata (L.) and the solitary wasp Cotesia rubecula (Marshall) (both Hymenoptera: Braconidae), whereas the main parasitoid of larval M. brassicae is the solitary endoparasitoid Microplitis mediator (Haliday) (Hymenoptera: Braconidae) (Harvey et al., 1999; Lauro et al., 2005). Moreover, eggs and larvae are preyed on by generalist predators such as spiders, chrysopids, staphylinids and carabids (Johansen, 1997). The aims of this 2-year study conducted at two farms in Switzerland were three-fold: (i) to assess whether predation rates on M. brassicae eggs can be enhanced by sown wildflower strips, (ii) to test whether the provision of wildflower strips increases parasitism rates in M. brassicae and P. rapae and (iii) to assess whether wildflower strips influence the spatial patterns in parasitism. 2. Material and methods
* Corresponding author. Tel.: +41 62 8657246; fax: +41 62 8657273. E-mail address: lukas.pfiffner@fibl.org (L. Pfiffner). 0167-8809/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.agee.2008.10.003
The experiments were conducted in north-western Switzerland at one organic farm in Bibern, canton Solothurn, and another one at
L. Pfiffner et al. / Agriculture, Ecosystems and Environment 129 (2009) 310–314
Murimoos, canton Bern. To assess the effects of sown wildflower strips on pest control, we monitored egg and larval parasitism as well as egg-predation rates in six cabbage fields. Each field was divided into two blocks, each block with a cabbage plot with and without sown wildflower strips. The plot size was 45 m 25 m. In 2004, two fields per site which bordered to the opposite side of the same wildflower strip were sampled, one field per site in 2005. The field margins of the control plots were covered by plastic sheets in order to prevent weeds from establishing. At Murimoos cabbage (Brassica oleracea var. capitata) was planted on 26 May at one side and on 8 June 2004 at the other side of the wildflower strip. In 2005 cabbage was planted at one side only on 20 June. At Bibern broccoli (Brassica oleracea var. silvestris) was planted on 3 May 2004 on one side of the wildflower strip, whereas cabbage was planted on the other side on 17 May. In 2005 cabbage was planted at one side only on 23 May. A 3 m 3 m grid design was used to analyse the spatial pattern in parasitism; 36 or 48 labelled plants per block were used for assessing egg parasitism (up to 18 m from the strip) and larval parasitism (up to 24 m from the strip), respectively. The mixture of the wildflower strip used in this study is officially recommended within the Swiss agro-environmental program. Wildflower strips were sown with a mixture of 24 wild flower species in spring or autumn preceding the investigations. The size of strips were 3 m 35 m. During the sampling periods six to 14 plant species were blooming (Table 1). Arable weeds occurred within the study fields at low level only, weed coverage ranging between 0.5% and 3%. No pesticides and mechanical control measures were applied during the experiments. Within each year cabbage was planted in new fields, with field size varying from 0.66 to 0.80 ha. 2.1. Assessing egg parasitism and predation To investigate parasitism and predation rates on eggs of M. brassicae, paper cards with egg batches originating from laboratory colonies were placed out in the fields. One egg batch each (average batch size 21 eggs) was pinned with a needle to the ground underneath a crop leaf at each of the 36 grid sampling points. Eggs were not older than 1 day and were placed out between 7–23 July 2004 and between 5–28 July 2005 at four dates each year. After a 3day-exposure period, all cards were collected and incubated in the laboratory for 4 weeks at 22 8C to rear egg parasitoids. Parasitism rates were calculated as the proportion of the number of parasitoids reared in relation to the number of eggs exposed. Egg predation rates were calculated by comparing the number of eggs per batch before and after field exposure; this was done in 2005 only. Using a binocular microscope, the number of eggs was counted and empty, damaged or missing eggs were considered as destroyed or removed by invertebrate predators. 2.2. Determining larval parasitism To determine the rates of hymenopteran parasitism in P. rapae and M. brassicae larvae a DNA-based approach was employed. Larvae of both lepidopteran species were collected at three dates between 20 July and 3 August 2005 (site Bibern: two dates only). At each date, all larvae found on 96 cabbage plants per treatment were collected; new plants were selected for each sampling date. Each larva was identified to species level, put in a 1.5 ml reaction tube and frozen at 28 8C for subsequent molecular analysis. The total DNA of each field-collected larva was extracted using the CTAB-protocol described in Juen and Traugott (2005). Small larvae were extracted as a whole, medium and large-sized larvae were cut in two or three pieces, which were subsequently
311
processed as separate samples. Within each batch of 30 samples, one negative extraction control was included to check for DNA carry-over contamination. Primer pairs S29/A27 and S31/A28 (Traugott et al., 2006) targeting C. glomerata and C. rubecula, respectively, were used in multiplex PCR for testing P. rapae larvae in polymerase chain reaction (PCR) for these two parasitoids (for reaction mix and cycling conditions see Traugott et al., 2006). To exclude false negative PCR results due to PCR inhibitors or DNA extraction failures, all DNA extracts of P. rapae which did not test positive for parasitoid DNA were tested with general primers (Folmer et al., 1994) to ensure their amplifyability. Each 20 ml PCR mix contained 200 mM dNTPs, 1 mM of each primer, 2 ml of 10x buffer, 3 mM MgCl2, 0.75 U Taq DNA polymerase (BioTherm, GeneCraft), 1 ml bovine serum albumin (20 mg/ml), and 2 ml of extracted DNA. The cycling conditions were 2 min at 94 8C, 30 cycles of 15 s at 94 8C, 15 s at 52 8C, 45 s at 72 8C and final elongation for 2 min at 72 8C. Based on these tests, eight larvae of P. rapae (n = 662) had to be excluded from the analysis. Mitochondrial cytochrome oxide subunit I sequences (GenBank accession numbers see Traugott et al., 2006) were used to design primer pairs S100/A100 (S100 (50 -30 ) TTTAGAATTAGGAATATCTGGGAA; A100 (50 –30 ) TAAACCTCTGTGTCCCAAAATT) and S121/A124 (S121 (50 –30 ) GAATTAGGAAACCCTGGATC; A124 (50 –30 ) AAAGAGTTAAAGAAGGGGGA) to amplify a 310 bp fragment from M. mediator and a 217 bp fragment from M. brassicae, respectively. The latter primer pair served as an internal control for amplification-failure in the multiplex PCR and lead to the exclusion of eight M. brassicae larvae (n = 153) from the analysis. Each 10 ml multiplex PCR mix contained 1.5 ml of extracted DNA, 1 multiplex PCR master mix (QIAGEN), 0.2 mM of each primer and 2.5 ml PCR-water. The cycling conditions were 15 min at 95 8C, 30 cycles of 30 s at 94 8C, 90 s at 60 8C, 60 s at 72 8C and final elongation for 10 min at 72 8C. The specificity of the diagnostic PCR assays was tested using the target parasitoid species, its host species the lepidopteran species Pieris brassicae L. and Plutella xylostella L. as well as Diadegma semiclausum Helle´n (Hymenoptera: Ichneumonidae; main parasitoid of P. xylostella), all being abundant species in cabbage fields. The PCR assays proved to be specific for the target species. All PCR assays included DNA of the target species as a positive control, DNA of the host as positive/negative control and PCR water allowing to detect DNA carry-over contamination. PCR products were loaded onto ethidium bromide-stained agarose gels, separated by size using electrophoresis and visualized on a UVtransiluminator. 2.3. Statistical analysis Mann–Whitney U-test was used for non-parametric data to test for differences between treatments in egg and larval parasitism rates on M. brassicae and P. rapae, and predation rates on M. brassicae eggs and pest densities using the pooled data of all sampling dates with the software SPSS version 10. Analyses were done separately with arcsin-transformed data for each field. Two blocks per field were used to test treatments wildflower vs. control. The analysis of the distance gradients from the wildflower strip into the field was performed using the pooled data of exposed eggs within the grid. Additionally, geostatistical analysis was applied as exploratory data analysis to supplement the statistical analysis described above and to generate hypotheses to be tested with non-spatial statistics. The inverse distance weighting (IDW) method was employed, providing a more robust interpolation algorithm than kriging and indicator kriging (Clark and Harper, 2000).
312
L. Pfiffner et al. / Agriculture, Ecosystems and Environment 129 (2009) 310–314
Fig. 3. Mean parasitism rates (S.E.M.) in caterpillars of Pieris rapae. Comparison of wildflower strip (+) and control () treatment at the two sites Bibern (B) and Murimoos (M) using Mann–Whitney U-test, *: significant difference, p < 0.05, ns: no significant difference, p > 0.05.
Fig. 1. Mean parasitism rates (S.E.M.) in Mamestra brassicae eggs in relation to the distance to the wildflower strip. Plots with (*) and without (*) wildflower strips at site Bibern (a) and Murimoos (b). Mann–Whitney U-test with significant difference (*) with p = 0.042.
3. Results In the 2004 experiments, placed-out sentinel eggs (n > 21,000) were parasitized by T. evanescens Westwood and Telenomus sp., the latter being significantly more abundant than T. evanescens. Overall egg parasitism rates did not differ between wildflower and control treatments nor between most distances to the wildflower strip (Fig. 1). At both sites, no influence of the treatment on the spatial pattern in egg parasitism could be found using geostatistical analysis. Only a general edge effect at Murimoos, with higher parasitism rates occurring along the field margin could be detected. In 2005, egg parasitism was very low (<1%) most likely due to low temperatures and rainy weather (data not shown), precluding any further analysis at both sites. Overall predation rates on M. brassicae eggs varied from 2.2% to 6.4%, with higher predation rates occurring close to field margins. At Bibern predation on M. brassicae eggs was significantly higher in plots adjacent to wildflower strips than in those bordering to control plots (Fig. 2), however, this pattern did not occur at Murimoos. Only lacewing larvae were observed to feed on egg batches in the field.
Fig. 2. Mean predation rates (S.E.M.) on 3-day-exposed eggs of Mamestra brassicae. Comparison of predation rates between wildflower strip (+) and control () treatment at the two sites Bibern (B) and Murimoos (M) using Mann–Whitney U-test, *: significant difference, p < 0.05, ns: no significant difference, p > 0.05.
Larval densities of both M. brassicae and P. rapae did not significantly differ between treatments (Mann–Whitney U-test with p > 0.05). Mean caterpillar densities per plant (S.E.M.) varied in M. brassicae from 1.1 0.16 (Bibern) to 1.6 0.07 (Murimoos) and in P. rapae from 1.4 0.09 (Bibern) to 3.2 0.15 (Murimoos). High larval parasitism rates between 45% and 75% were detected in M. brassicae at both sites and in P. rapae at Murimoos. At Bibern, however, only one out of 106 collected P. rapae larvae was positive for parasitoid DNA (Fig. 3). At Murimoos, C. rubecula was the main parasitoid of P. rapae (six out of 548 caterpillars (1.09%) with another four yielding DNA of both Cotesia species). At Murimoos P. rapae larvae showed significantly higher parasitism rates in plots bordering wildflower strips than in control plots, whereas no such effect could be found at Bibern (Fig. 3). Parasitism rates in M. brassicae larvae did not differ significantly between treatments at both sites. 4. Discussion The provision of wildflower strips in organic cabbage cultures provided an idiosyncratic effect on the control of lepidopteran cabbage pests: parasitism rates of M. brassicae eggs and larvae as well as of P. rapae larvae at Bibern were not affected, whereas parasitism rates of larval P. rapae at Murimoos and predation rates on M. brassicae eggs were significantly enhanced by the wildflower strips. Only a few field studies have assessed the impact of flower strips on pest control in vegetable crops and their results are contradictory. Lavandero et al. (2005), Lee and Heimpel (2005), Winkler (2005) and Winkler et al. (2006) have shown that flower strips can increase parasitism rates in neighbouring fields, but there are also studies reporting on no differences in parasitism rates between fields with and without flower strips (e.g. Berndt et al., 2002). The ability of parasitoids to utilize floral resource subsidies is dependent on a wide range of factors, including flower morphology, colour, odour or the timing of nectar production (Heimpel and Jerwis, 2005). Considering all these aspects, only a few of the plant species used in the seed mixture such as C. cyanus, F. esculentum and D. carota may have benefitted the target parasitoids (Lee and Heimpel, 2005; Winkler, 2005). Although parasitism rates in the artificially exposed M. brassicae eggs ranged from 14% to 52%, the wildflower strips did not enhance egg parasitism rates. Moreover, parasitism rates did not change significantly with increasing distance to the field border in both plots with and without wildflower strips. Egg parasitoids cover only short distances by active dispersal (Fournier and Boivin, 2000)
L. Pfiffner et al. / Agriculture, Ecosystems and Environment 129 (2009) 310–314 Table 1 Seed mixture of wildflower strip and occurrence of blooming wild flowers during the sampling period in 2004 and 2005 at both sites. Sown species
Blooming plant species g/ha
Achillea millefolium L. Agrostemma githago L. Anthemis tinctoria L. Centaurea cyanus L. Centaurea jacea L. Cichorium intybus L. Daucus carota L. Dipsacus silvester L. Echium vulgare L. Fagopyrum esculentum Miller Hypericum perforatum L. Leucanthemum vulgare LAM. Legousia speculum-veneris L. Malva moschata L. Malva sylvestris L. Melilotus albus MED. Onobrychis viciifolia Scop. Origanum vulgare L. Papaver rhoeas L. Pastinaca sativa L. Silene alba L. Tanacetum vulgare L. Verbascum densiflorum L. Verbascum lychnitis L.
20 600 20 500 200 120 150 2 200 7245
Unsown dominant species Matricaria chamomilla L. Total number
Bibern 2004
Murimoos
x x x x
x x x
x
Bibern 2005
Murimoos
x x x x x x x r x r
r x x x
60 80
x
r x
30
r
r
20 60 20 600 60 150 80 100 3 50 30
x
r
x
r x
r
x
r
x
x
x
6
14
10
x 6
parasitism rates in M. brassicae eggs correlated positively with the area of woody habitats and pasture at the field level. Besides attracting predators and parasitoids, introducing nectar sources to a cropping system could also increase the density of pest insects which might benefit from the same food sources (Baggen et al., 1999; Zhao et al., 1992). No such effect was found in the present study: larval densities of M. brassicae and P. rapae were similar in plots with and without wild flower strips. In conclusion, our results indicate that the floral treatment using a multiple species mixture did not consistently improve the control of P. rapae and M. brassicae compared with a nearby control treatment under the small-scaled vegetable crop production situation in Switzerland. A more effective conservation biological control may be achieved by the provision of selective floral resources (Lavandero et al., 2006) and the creation of permanent non-crop habitats as e.g. woodlots and hedgerow (Den Belder et al., 2006). Acknowledgements
x
0
313
r: rare plant sp.; x > 1% cover.
but are able to recolonise from overwintering habitats such as pastures (Schneider et al., 2003). The predation on M. brassicae eggs was found to be higher on plants near to the wildflower strip. Wildflower strips are known to harbour many epigeic polyphagous arthropods during winter (Pfiffner and Luka, 2000), which can immigrate during spring into the adjacent arable crops (Holland et al., 2003), causing the high predation rates observed. Parasitism rates in larval M. brassicae were not affected by the wildflower strip, whereas in P. rapae larvae significantly higher parasitism rates were observed at one of the two sites. The latter finding is in accordance with the study of Lee and Heimpel (2005), finding that buckwheat (F. esculentum) strips increased parasitism rates in caterpillars of P. rapae. Both C. glomerata and C. rubecula are known to parasitize P. rapae; however, C. rubecula regularly outcompetes C. glomerata in this host (Geervliet et al., 2000). This was confirmed by our study as 0.7% of the caterpillars were parasitized by both Cotesia species; just 0.4% of them were parasitized by C. glomerata only. In a previous study using the same PCR assay (Traugott et al., 2006), no multiparasitism was detected in P. rapae and P. brassicae larvae collected at Murimoos in 2004, yet a total of only 157 larvae was screened. Interestingly, parasitism rates in P. rapae larvae were quite similar in 2004 (53.4%, n = 30) and in 2005 (53.4%, n = 548) at this site, suggesting that site-specific effects on larval parasitism occur. Permanent non-crop habitats may play a crucial role for parasitoids to build up their populations (Thies and Tscharnke, 1999). For example, the presence of woodlots and hedgerows can negatively affect the number of P. rapae larvae in arable land (Den Belder et al., 2006). Bianchi et al. (2005) found that predation and
This study was financially supported by Federal Office for the Environment, Federal Office for Agriculture, Fonds Landschaft Schweiz, Canton of Fribourg and Bern and MAVA Foundation, Sophie-Carl Binding Foundation. We thank H. Baur, K. van Achterberg K. Horstmann and J. C. Monje for the support in parasitoid determination. The authors are grateful to the farmers for providing the fields as well as for the comments on the manuscript provided by anonymous reviewers. References Baggen, L.R., Gurr, G.M., Meats, A., 1999. Flowers in tri-trophic systems: Mechanisms allowing selective exploitation by insect natural enemies for conservation biological control. Entomol. Exp. Appl. 91, 155–161. Berndt, L.A., Wratten, S.D., Hassan, P.G., 2002. Effects of buckwheat flowers on leafroller (Lepidoptera: Tortricidae) parasitoids in New Zealand vineyard. Agric. Forest Entomol. 4, 39–45. Bianchi, F.J.J.A., van Wingerden, W.K., Griffioen, A.J., van der Veen, M., van der Straten, M.J., Wegman, R.M.A., Meeuwsen, H.A.M., 2005. Landscape factors affecting the control of Mamestra brassicae by natural enemies in Brussels sprout. Agric. Ecosyst. Environ. 107, 145–150. Clark, I., Harper, W.V., 2000. Practical Geostatistics. Ecosse North America Ltd., Columbus, OH. Den Belder, J., Elderson, J., Schelling, G., 2006. Landscape effect on the abundance of lepidopteran pests in Brussel sprouts. IOBC/wprs Bulletin 29, 25–28. Folmer, O., Black, M., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Marine Biol. Biotech. 3, 294–299. Fournier, F., Boivin, G., 2000. Comparative dispersal of Trichogramma evanescens and Trichogramma pretiosum (Hymenoptera: Trichogrammatidae) in relation to environmental conditions. Popul. Ecol. 29, 55–63. Geervliet, J.B.F., Verdel, M.S.W., Snellen, H., Schaub, J., Dicke, M., Vet, L.E.M., 2000. Coexistence and niche segregation by Weld populations of the parasitoids Cotesia glomerata and C. rubecula in the Netherlands: predicting Weld performance from laboratory data. Oecologia 124, 55–63. Gurr, G.M., Wratten, S.D., Altieri, M.E., 2004. Ecological Engineering for Pest Management: Advances in Habitat Manipulation for Arthropods. CSIRO Publishing, Collingwood, Vic, Australia. Harvey, J.A., Jervis, M.A., Gols, R., Jiang, N., Lem, M.V., 1999. Development of the parasitoid Cotesia rubecula (Hymeroptera: Braconidae) in Pieris rapae and Pieris brassicae (Lepidoptera: Pieridae): evidence for host regulation. J. Ins. Physiol. 45, 173–182. Heimpel, G.E., Jerwis, M.A., 2005. Does nectar improve biocontrol by parasitoids? In: Wa¨ckers, F., van Rijn, P.C.,Bruin, J. (Eds.), Plant provided food for carnivorous insects: a protective Mutalism and its Applications. Chapman & Hall, London, UK, pp. 267–304.
Plant provided food for carnivorous insects ¨ komassnahmen, Bereich BiodiHerzog, F., Walter, T. (Eds.), 2005. Evaluation der O versita¨t, 56. Schriftenreihe FAL, pp 205. Hill, D.S., 1987. Agricultural Insect Pests of Temperate Regions and their Control. Cambridge Univ. Press, Cambridge.
314
L. Pfiffner et al. / Agriculture, Ecosystems and Environment 129 (2009) 310–314
Holland, J., Birkett, T., Begbie, M., Southway, S., Thomas, C.F.G., 2003. The spatial dynamics of predatory arhropods and the importance of crop and adjacent margin habitats. IOBC/wprs Bull. 26, 65–70. Jervis, M.A., Heimpel, G.E., 2005. Phytophagy. In: Jervis, M.A. (Ed.), Insects as natural enemies. A practical perspective. Springer, Dordrecht, pp. 525–555. Johansen, N.S., 1997. Mortality of eggs, larvae and pupae and larval dispersal of the cabbage moth, Mamestra brassicae, in white cabbage in south-eastern Norway. Entomol. Exp. Appl. 83, 347–360. Juen, A., Traugott, M., 2005. Detecting predation and scavenging by DNA gutcontent analysis: a case study using a soil insect predator–prey system. Oecologia 142, 344–352. Landis, D.A., Wratten, S.D., Gurr, G.M., 2000. Habitat management to conserve natural enemies of arthropod pests in agriculture. Ann. Rev. Entomol. 45, 175–201. Lauro, N., Kuhlmann, U., Mason, P.G., Holliday, N.J., 2005. Interaction of a solitary larval endoparasitoid, Microplitis mediator, with its host. Mamestra brassicae: host acceptance and host suitability. J. Appl. Entomol. 129, 567–573. Lavandero, B., Wratten, S., Didham, R., Gurr, G.M., 2006. Increasing floral diversity for selective enhancement of biological control agents: a double-edged sward? Basic Appl. Ecol. 7, 236–243. Lavandero, B., Wratten, S., Shishehbor, P., Worner, S., 2005. Enhancing the effectiveness of the parasitoid Diadegma semiclausum (Hellen): movement after use of nectar in the field. Biol. Control 34, 152–158. Lee, J.C., Heimpel, G.E., 2005. Impact of flowering buckwheat on lepidopteran pests and their parasitoids at two spatial scales. Biol. Control 34, 290–301.
Pfiffner, L., Luka, H., 2000. Overwintering of arthropods in soils of arable fields and adjacent semi-natural habitats. Agric. Ecosyst. Environ. 78, 215–222. Pfiffner, L., Wyss, E., 2004. Use of sown wildflower strips to enhance natural enemies of agricultural pests. In: Gurr, G., Wratten, S., Altieri, M. (Eds.), Ecological Engineering for Pest Management: Advances in Habitat Manipulation for Arthropods. CSIRO Publishing, Collingwood, Vic, Australia, pp. 165– 186. Schneider, C., Dover, J., Fry, G.L.A., 2003. Movement of two grassland butterflies in the same habitat network: the role of adult resources and size of the study area. Ecol. Entomol. 28, 219–227. Thies, C., Tscharnke, T., 1999. Landscape structure and biological control in agroecosystems. Science 6, 893–895. Traugott, M., Zangerl, P., Juen, A., Schallhart, N., Pfiffner, L., 2006. Detecting key parasitoids of lepidopteran pests by multiplex PCR. Biol. Control 39, 39–46. Winkler, K., 2005. Assessing the risks and benefits of flowering field edges. Strategic use of nectar sources to boost biological control. Thesis, Wageningen University, p. 118. Winkler, K., Wa¨ckers, F., Bukovinszkine-Kiss, G., van Lenteren, J.C., 2006. Sugar resources are vital for Diadegma semiclausum fecundity under field conditions. Basic Appl. Ecol. 7, 133–144. Zhao, J.Z., Ayers, G.S., Grafius, E.J., Stehr, F.W., 1992. Effects of neighboring nectarproducing plants on populations of pest lepidoptera and their parasitoids in broccoli plantings. The Great Lakes Entomologist 25, 253–258.