IOBC / WPRS

Working Group „GMOs in Integrated Plant Production“

Proceedings of the third meeting on

Ecological Impact of Genetically Modified Organisms

at

Warsaw (Poland)

23 – 25 May, 2007

Editors:

Jörg Romeis, Michael Meissle & Olivier Sanvido

IOBC wprs Bulletin Bulletin OILB srop Vol. 33, 2008

The content of the contributions is in the responsibility of the authors

The IOBC/WPRS Bulletin is published by the International Organization for Biological and Integrated Control of Noxious Animals and Plants, West Palearctic Regional Section (IOBC/WPRS) Le Bulletin OILB/SROP est publié par l‘Organisation Internationale de Lutte Biologique et Intégrée contre les Animaux et les Plantes Nuisibles, section Regionale Ouest Paléarctique (OILB/SROP) Copyright: IOBC/WPRS 2008

The Publication Commission of the IOBC/WPRS: Horst Bathon Julius Kühn Institute (JKI), Federal Research Centre for Cultivated Plants Institute for Biological Control Heinrichstr. 243 D-64287 Darmstadt (Germany) Tel +49 6151 407-225, Fax +49 6151 407-290 e-mail: horst.bathon@jki.bund.de

Luc Tirry University of Gent Laboratory of Agrozoology Department of Crop Protection Coupure Links 653 B-9000 Gent (Belgium) Tel +32-9-2646152, Fax +32-9-2646239 e-mail: luc.tirry@ugent.be

Address General Secretariat: Dr. Philippe C. Nicot INRA – Unité de Pathologie Végétale Domaine St Maurice - B.P. 94 F-84143 Montfavet Cedex (France) ISBN 978-92-9067-207-4 http://www.iobc-wprs.org

Table of Contents Preface ....................................................................................................................................... i Contents ................................................................................................................................... iii List of Participants ................................................................................................................. vii I. Key notes Exploring the potential of corn borers to develop resistance to Bt-corn in Europe.

Juan Ferré, Joel González-Cabrera, Yolanda Bel, Baltasar Escriche ............................ 1 The impacts of novel management on ecosystem dynamics; tales from the UK Farm

Scale Evaluations of GMHT crops. Alison J. Haughton, David A. Bohan ................................................................................ 7

Integrating insect-resistant GM crops in pest management systems. Steven E. Naranjo ........................................................................................................... 15

II. Presentations Impact of glyphosate use on arthropods in transgenic herbicide-tolerant maize;

preliminary results from studies in Spain. Ramon Albajes, Matilde Eizaguirre, Daniel Casado, Meritxell Pérez, Carmen López, Belén Lumbierres, Xavier Pons........................................................................... 23

Preventing spread of Ostrinia nubilalis Hbn. by cultivation of Bt transgenic maize – First field experiments in southeastern Poland Paweł K. Bereś, Robert Gabarkiewicz............................................................................ 31

Baseline susceptibility of Helicoverpa armigera (Hübner) to Bt toxins Cry1Ac and Cry2Ab2 in West Africa. Thierry Brevault, Patrick Prudent, Maurice Vaissayre.................................................. 37

Direct effects of Galanthus nivalis agglutinin (GNA) and avidin on the ladybird beetle Coccinella septempunctata. Mukesh K. Dhillon, Nora C. Lawo, H.C. Sharma, Jörg Romeis .................................... 43

Validation of some techniques used in the evaluation of GM plant effects on tri-trophic interactions. Julia Górecka, Zbigniew T. Dąbrowski, Monika Godzina, Karolina Kubis................... 51

Round robin quantitation of Cry3Bb1 using the qualitative PathoScreen ELISA. Hang Thu Nguyen, Heinz Hunfeld, Michael Meissle, Rona Miethling-Graff, Sibylle Pagel-Wieder, Stefan Rauschen, Corinne Zurbruegg, Sabine Eber, Frank Gessler, Jörg Romeis, Christoph C. Tebbe, Wolfgang Nentwig, Johannes A. Jehle ...... 59

F2 Screen and field sampling with light trap cages, two methods for a resistance moni-toring in Bt crops. Heike Engels, Ingolf Schuphan, Sabine Eber ................................................................. 67

Diversity and seasonal phenology of spiders, ground beetles and rove beetles in conventional and transgenic maize in Central Spain. Gema P. Farinós, Marta de la Poza, Pedro Hernández-Crespo, Félix Ortego, Pedro Castañera ............................................................................................................. 75

Changes in biochemistry of cucumber carrying the thaumatin II gene: relevance to herbivores. Małgorzata Kiełkiewicz, Janina Gajc-Wolska, Maria Szwacka, Stefan Malepszy......... 79

Belowground volatile emission of Bt maize after induction of plant defence. Michael Meissle, Ivan Hiltpold, Ted C. J. Turlings, Jörg Romeis.................................. 85

Assessment of possible non-target impacts of the novel Bt-maize event MON88017 resistant against the Western Corn Rootworm Diabrotica virgifera virgifera (LeConte). Stefan Rauschen, Ingolf Schuphan, Sabine Eber ............................................................ 93

Application of environmental risk assessments of pest resistant crops in different environments. Jeremy B. Sweet ............................................................................................................ 101

Ground beetles (Coleoptera: Carabidae) in transgenic herbicide tolerant maize hybrids: Impact of the transgenic crop or the weed control practice? Dóra Szekeres, Ferenc Kádár, Zita Dorner.................................................................. 105

Reduction of mycotoxin threats to mammals and birds through the cultivation of Bt maize cultivars in Poland. Agata Tekiela, Robert Gabarkiewicz ............................................................................ 111

Can plants produced from callus culture be used as near-isogenic standards in comparative analyses of transgenic potato clones? Ramona Thieme, Helmut Griess, Thomas Thieme........................................................ 117

A method for selecting non-target organisms for testing the biosafety of GM plants. Jacqui H. Todd, Padmaja Ramankutty, Louise A. Malone ........................................... 123

Impact of transgenic Bt corn on European corn borer (Ostrinia nubilalis Hübner) in Lower Silesia, Poland. Preliminary results. J.P. Twardowski, M. Hurej, L. Kordas ......................................................................... 129

Non-target organism risk assessment in Bt crops. Zigfridas Vaituzis .......................................................................................................... 133

Transgenic Escherichia coli co-expressing cry1Ca and cry1Ac: toxicity and synergy against three agricultural pests. Arieh Zaritsky, Eitan Ben-Dov...................................................................................... 139

Assessing the effects of Bt-maize pollen on Typhlodromus pyri (Acari: Phytoseiidae). Rostislav Zemek, Zuzana Vávrová ................................................................................ 145

III. Report from a special WG activity Non-target arthropod risk assessment of insect-resistant GM crops.

Jörg Romeis .................................................................................................................. 149 Special activity: Non-target risk assessment and regulation. Protocol of the discussion.

Elisabeth Schulte........................................................................................................... 157

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 1-6 Exploring the potential of corn borers to develop resistance to Bt-corn in Europe Juan Ferré, Joel González-Cabrera, Yolanda Bel, Baltasar Escriche Fac. de Ciències Biològiques, Universitat de València, Dr. Moliner 50, 46100-Burjassot (València), Spain (E-mail: Juan.Ferre@uv.es) Abstract: Bacillus thuringiensis genes coding for insecticidal proteins (cry genes or Bt genes) have been transferred to agronomically important crops rendering the so called Bt-crops, which are protected from insect attack. Although no case of field resistance to Bt-crops has been reported so far, the potential of insects to evolve resistance to insecticides is well known, and it is considered to be one of the threats posed to genetically modified insecticidal plants. It is for this reason that the EU funded a project involving 11 groups to evaluate the potential of European corn borer populations to develop resistance to the currently deployed varieties of Bt-corn, the only transgenic insecticidal crop allowed for commercial planting in the EU. The objectives of our group in this project were, among others, to determine the biochemical basis of the mode of action of B. thuringiensis toxins in corn borers, and to characterise the cadherin-like gene in O. nubilalis, a major candidate resistance gene, for the future application to molecular monitoring of resistance alleles. Key words: Cry proteins, Ostrinia nubilalis, Sesamia nonagrioides, Bacillus thuringiensis, European corn borer, Mediterranean corn borer, binding sites, mechanisms of resistance, cadherin, resistance genes Introduction After 10 years of Bt-corn planting in the US, no development of resistance in corn borer populations to Bacillus thuringiensis (Bt) insecticidal proteins has been found in the field. One reason is for sure the strict resistance management imposed by the US Environmental Protection Agency (EPA) to the seed companies: a high expression of the Bt protein in the plant and the use of refuges planted with non-transformed plants. Europe has started with the adoption of Bt-corn much more cautiously and most of the countries have been reluctant to approve Bt-corn planting. Spain has been the exception, and it has been always the European country with the largest area planted to Bt-corn, reaching over 60,000 ha in 2006 (James, 2006).

One of the main threats of adopting Bt-corn is the high selection pressure imposed to the corn borer populations which can lead to the development of resistance. It was for this reason that the EU funded a project with the title “Protecting the benefits of Bt-toxins from insect resistance development by monitoring and management (ProBenBt)” (contract number QLK3-CT-2002-01969), which covered the period from November 2002 till April 2006. The project was coordinated by Ingolf Schuphan from Aachen University (Aachen, Germany) and involved 11 research groups. The objectives of this project were, among others, to characterise the corn borer populations in Europe to determine the genetic diversity and the frequency of resistance alleles.

In the present paper we will summarize the results of our group, which was involved in the mode of action of Bt insecticidal proteins in the corn borers Ostrinia nubilalis (Hübner) (Lepidoptera: Crambidae) and Sesamia nonagrioides (Lefebvre) (Lepidoptera Noctuidae), the

1

2

biochemical characterisation of potential mechanisms of resistance, and the characterisation of one of the major candidate genes (cadherin-like) for the future application to molecular monitoring of resistance alleles.

Materials and Methods Binding studies Binding studies were performed using 125I-labeled or biotin-labeled B. thuringiensis toxins and brush-border membrane vesicles (BBMV) as described elsewhere (Li et al., 2004; González-Cabrera et al., 2006). We used the toxins that had been previously shown to be active against each insect species. These were Cry1Aa, Cry1Ab, Cry1Ac, and Cry1Fa for both species and Cry1Ca for S. nonagrioides. The toxins were purified from B. thuringiensis strains producing a single crystal protein, and they were activated with commercial bovine trypsin. By competing the labeled toxin with the same unlabeled toxin (homologous competition) for binding to the BBMV, the binding affinity (Kd) and the concentration of binding sites (RRt) were calculated. By competing the labeled toxin with different unlabeled toxins (heterologous competition), we obtained information on whether different toxins bind to the same binding site. Gene structure characterisation Genomic DNA was obtained from the thoraxes of frozen adult insects using the DNeasy Tissue Kit (Quiagen GmbH, Hilden, Germany). Gene fragments were amplified using primers based on the reported cDNA sequence AX147201 (Flannagan et al., 2001).

PCR was performed using the Expand High Fidelity PCR System (Roche, Mannheim, Germany). The reaction products were purified (using the High Pure PCR Product Purification Kit, Roche) and sequenced. When necessary, amplicons were cloned into the pGEM-T cloning system (Promega, Madison, WI, USA) and transformed into competent E. coli XL1-Blue cells.

The analyses and assembly of the sequences was performed using the SequencherTM 4.0.5. software (Gene Codes Corp., Ann Arbor, MI, USA). For database searches and feature analyses the following programs were used: BLAST program at the National Centre for Biotechnology Information, Tandem Repeats Finder Program ver. 2.02, EMBOSS:cpgplot utility, Repeat Masker ver. 1.165, UTRScan and UTRBlast at UTResource, TFDSearch, the Neural Network Promoter Prediction program, and the WWW Promoter Scan Program). For the in silico analysis of the predicted protein, we used the CLUSTAL X ver. 1.83 software, the BioEdit ver. 7.0.0 program, the SMART and IntrProScan tools and the Cadherin Resource Database. Results and Discussion Binding site model in O. nubilalis and S. nonagrioides and its use at predicting cross-resistance Using different toxins and competition experiments, a binding site model has been proposed for each of the corn borer species studied. To illustrate this point, Figure 1 shows the results obtained with BBMV from O. nubilalis using 125I-labeled Cry1Ab and, as competitors, the four Cry1 toxins that are active against this insect. Using different labeled toxins and competitors, in both insect species, Cry1A toxins competed among each other and with Cry1Fa for the same binding site, although Cry1Fa competed with very low affinity (Figures 1, 2 and 3). Cry1Fa was not labelled for the binding assays in O. nubilalis, therefore, we could

3

not obtain any information for a binding site other than that shared with Cry1A toxins (Figures 1 and 2). However, in the experiments carried out with S. nonagroides BBMV, we labeled Cry1Fa with biotin and performed competition experiments with unlabeled Cry1Ab. This revealed the occurrence of a high affinity binding site for Cry1Fa in this species (Figure 3) (González-Cabrera et al., 2006). Cry1Ca was found to have a unique binding site and did not compete for Cry1A’s binding site in S. nonagrioides (Figure 3). Competition between Cry1Ca and Cry1Fa binding sites was not tested.

00

20

40

60

80

100

120

10-2 10-1 100 101 102 103

Cry1Ab (Homologous)

Cry1Fa

Cry1Aa

Cry1Ac

Competitor (nM)

125 I-C

ry1A

b in

itial

bin

ding

(%)

Figure 1. Binding of 125I-labeled Cry1Ab to BBMV from O. nubilalis in the presence of increasing concentrations of unlabeled competitor. Cry1Ab Cry1AcCry1Aa Cry1Fa

?

Figure 2. Binding site model for Cry toxins binding to BBMV from O. nubilalis. The width of the arrows represents relative binding affinities.

The binding site models have an important predictive value regarding cross-resistance,

especially when resistant strains are not yet available. An alteration of a common binding site could affect binding of the toxins that use it for attaching to the membrane and, consequently, this could confer resistance against all of them. Therefore, sharing a binding site indicates

4

potential for development of cross-resistance by the mechanism of binding site alteration. However, it must be kept in mind that there are other mechanisms of resistance which do not depend on the modification of the binding site, although these usually confer only low levels of resistance (Ferré & Van Rie, 2002).

Our results indicate that combination of Cry1A toxins in a same corn plant would not be useful for resistance management. However, combination of a Cry1A toxin with Cry1Ca would confer protection against resistance development in S. nonagrioides (O. nubilalis is not susceptible to Cry1Ca). Prediction of the outcome of the combination of Cry1A with Cry1Fa may be more problematic, since these toxins share a common receptor. However, the fact that Cry1Fa only binds with low affinity the Cry1A high affinity binding site, along with the results obtained with S. nonagrioides showing a high affinity binding site for Cry1Fa different from the Cry1A high affinity binding site, seems to point at different targets of these toxins to exert their toxic action.

Cry1Fa Cry1A’s Cry1Ca

Figure 3. Binding site model for Cry toxins binding to BBMV from S. nonagrioides. The width of the arrows represents relative binding affinities.

Genomic structure of the gene coding for the cadherin membrane protein in O. nubilalis The genomic sequence obtained for the cadherin-like gene in O. nubilalis encompassed 19.6 kb (GenBank accession no. DQ000165), that is, about four times the size of the reported cDNA sequence (5.5 kb). The comparison of the genomic and cDNA sequences showed a genomic structure with 34 introns, one of them in the 5’-UTR region (Figure 4). All the intron-exon boundaries followed the AG-GT rule. The possibility of alternative promoters in the 5’-UTR region (including intronic and exonic sequences) was tested with prediction programs based on eukaryotic promoter sequences for RNA polymerase II and on specific Drosophila melanogaster promoter sequences. Only in this last case, 3 positions (-170, -839 and -868) were recognized as possible promoter regions inside intron -1. The in silico examination of regulatory or significant features, such as internal ribosome entry sites, did not render any consistent sequence. Two kinds of short tandem repeats (STR) were detected (one in the intron 25 and the other one covering the exon and intron 31), and no traces of retroelements or DNA transposons were observed either in exons or introns (Bel and Escriche, 2006).

5

Figure 4. Intron-exon structure of the O. nubilalis cadherin-like gene. The exons are represented as black rectangles.

The cadherin protein sequence inferred from the cadherin-like gene consisted of 1717 aa, which is in agreement with other authors (Coates et al., 2005; Flannagan et al., 2005). The protein shared about 60% of identity with other cadherin-like proteins reported for other Lepidoptera, and, in addition, it exhibited a common pattern of secondary structures with them (Pigott & Ellar, 2007): an extracellular part with 22-aa signal peptide and 12 cadherin repeats (about 100 aa each one), a 23-aa transmembrane domain, and 123-aa intracellular domain. The analysis of this protein using the Cadherin Resource Database suggested the grouping of the sequence into the protocadherin family, with unknown function.

The sequence variability in some regions of this gene could be a useful tool to design molecular probes for the monitoring of resistance alleles in this species and the unravelling of the biological function of this protein can help in the resistance management strategies. Acknowledgments This work was funded by a grant from the European Union (ref. QLRT-2001-01969).

References Bel, Y. & Escriche, B. 2006: Common genomic structure for the Lepidoptera cadherin-like

genes. Gene 381: 71-80. Coates, B.S., Sumerford, D.V., Hellmich, R.L. & Lewis, L.C. 2005: Sequence variation in the

cadherin gene of Ostrinia nubilalis: a tool for field monitoring. Insect Biochem. Mol. Biol. 35: 129-139.

Ferré, J., & Van Rie, J. 2002: Biochemistry and genetics of insect resistance to Bacillus thuringiensis. Annu. Rev. Entomol. 47: 501-533.

Flannagan, R.D., Mathis, J.P. & Meyer, T.E. 2001: Novel Bt toxin receptors from lepidopteran insects and methods of use. World Intellectual Property Organization, Patent WO 01/36639 A2.

Flannagan, R.D., Yu, C.-G., Mathis, J.P., Meyer, T.E., Shi, X., Siqueira, H.A.A. & Siegfried, B.D. 2005: Identification, cloning and expression of a Cry1Ab cadherin receptor from European corn borer, Ostrinia nubilalis (Hübner) (Lepidoptera: Crambidae). Insect Biochem. Mol. Biol. 35: 33-40.

James, C. 2006: Executive summary of global status of commercialized biotech/GM crops: 2006. ISAAA Brief No. 35. International Service for the Acquisition of Agri-Biotech Applications, Ithaca, NY.

González-Cabrera, J., Farinós, G.P., Caccia, S., Díaz-Mendoza, M., Castañera, P., Leonardi, M.G., Giordana, B. & Ferré, J. 2006: Toxicity and mode of action of Bacillus thuringiensis Cry proteins in the Mediterranean corn borer, Sesamia nonagrioides (Lefebvre). Appl. Environ. Microbiol. 72: 2594-2600.

6

Li, H.R., González-Cabrera, J., Oppert, B., Ferré, J., Higgins, R.A., Buschman, L.L., Radke, G.A., Zhu, K.Y. & Huang, F.N. 2004: Binding analyses of Cry1Ab and Cry1Ac with membrane vesicles from Bacillus thuringiensis-resistant and -susceptible Ostrinia nubilalis. Biochem. Bioph. Res. Co. 323: 52-57.

Piggot, C.R.& Ellar, D.J., 2007: Role of receptors in Bacillus thuringiensis crystal toxin activity. Microbiol. Mol. Biol. Rev. 71: 255-281.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 7-13

The impacts of novel management on ecosystem dynamics; tales from the UK Farm Scale Evaluations of GMHT crops Alison J. Haughton, David A. Bohan Ecosystem Dynamics & Biodiversity Group, Department of Plant and Invertebrate Ecology, Rothamsted Research, West Common, Harpenden, Herts, AL5 2JQ. UK (E-mail: alison.haughton@bbsrc.ac.uk). Abstract: Concerns about the possible negative impact of the novel herbicide management associated with GM herbicide-tolerant (GMHT) crops on British farm wildlife led to the establishment of the Farm Scale Evaluations (FSEs). This series of field trials evaluated wildlife changes by comparing the wildlife of a GMHT crop, with its associated herbicide management, against a conventional variety and current ‘best practice’ herbicide management. Using a half-field design, in some 65 fields per GMHT crop distributed across the arable growing areas of Great Britain, the abundance, biomass and diversity of weed plant and invertebrate species or taxa were assessed. The results showed that there were marked changes in some groups of weed plants and invertebrates with GMHT management. A basic assumption for the trials was that there were no direct effects of GMHT herbicide management on the invertebrates, and that any effects on invertebrates were caused by changes in the weed plants due to differences in herbicide management. Although the experimental approach taken was scientifically rigorous, the repeated testing of the null hypothesis for each species or taxa was bound to generate significant effects just by chance. What was not clear was whether all the observed changes indicated a broad risk to wildlife in farmland, and were the changes important? The aim of our research in the Ecosystem Dynamics and Biodiversity group is to understand whether there are better approaches to understanding changes in ecosystems with management, and the risk posed. Using analyses of data from the FSEs, we outline some hypotheses for agro-ecosystem structuring, the impact of management on this, and the wider risks to wildlife of these changes. Specifically, we ask: can we detect changes in the agro-ecosystem with novel management; how are invertebrates linked to plants, and can we observe effects of novel management on these links; and can we extrapolate to biodiversity groups of social importance, such as birds, but which are difficult to measure at the field-scale? Key words: GM, herbicide tolerant, biodiversity, multivariate analysis, trophic relationships, birds Introduction Concerns about the possible negative impact of the novel herbicide management associated with GM herbicide-tolerant (GMHT) crops on British farm wildlife led to the establishment of the Farm Scale Evaluations (FSEs). This series of field trials evaluated wildlife changes by comparing the wildlife of a GM crop, with its associated herbicide management, against a conventional variety and current ‘best practice’ herbicide management. Using a half-field design, in some 65 fields per GMHT crop distributed across the arable growing areas of Great Britain, the abundance, biomass and diversity of weed plant and invertebrate species or taxa was assessed.

The results showed that there were marked changes in some groups of weed plants and invertebrates with GMHT management. A basic assumption for the trials was that there were no direct effects of GMHT herbicide management on the invertebrates, and that any effects on invertebrates were caused by changes in the weed plants due to differences in herbicide

7

8

management. Although the experimental approach taken was scientifically rigorous, the repeated testing of the null hypothesis for each species or taxa was bound to generate significant effects just by chance. What was not clear was whether all the observed changes indicated a broad risk to wildlife in farmland?

Results and Discussion In the Ecosystem Dynamics and Biodiversity Group (EDBG) at Rothamsted Research, we hypothesised that for an effect of GMHT to be an important risk to wildlife in farmland, significant shifts in the communities of weeds and invertebrates present in the FSE fields would have to be apparent. We therefore adopted Canonical Variates Analysis (CVA), which is a multivariate technique, to test for compositions of farmland communities in different crops and in the conventional and GM field halves. Counts of weed species and invertebrate taxa were converted to compositional proportions. We found that a significant proportion of the variation in the plant and invertebrates was due to the crop sown, and in the case of the weeds to the herbicide management used. In effect, we found the community of weeds changed with the crop sown and the herbicide management used (Figure 1a) and that the community of invertebrates changes with the crop grown, but not with herbicide management (Figure 1b) (Smith et al., submitted). This would suggest, then, the applied management determines the observed weed and invertebrate communities of fields. It would also indicate that there is no single, common weed/invert community across farmland. For risk assessments, though, our findings would indicate that GMHT cropping would be an important risk to weed biodiversity, but that there would be no important impact on the invertebrates.

a) b)

Figure 1. CVA scores (group means), with 95% confidence regions, plotted pairwise for the first two dimensions for: a) the weed species compositions; and b) invertebrate taxa, for spring-sown beet (B), maize (M), oilseed rape (S) and winter-sown oilseed rape (W), under GMHT (shaded and subscript GM) and conventional (open and subscript C) herbicide management. The percentage variation explained by each dimension is presented in parentheses within the axis titles.

9

These analyses indicate that management can strongly affect the composition of weeds and invertebrates we observe in farmland, but the work does not show whether this modifies ecological functioning. Here we extend our community approaches to a functional group description in order to: study services provided to the ecosystem; extrapolate our findings to higher trophic levels, such as birds; and provide a generic basis for modelling ecosystems. Here we highlight two examples using food for birds and trophic groups.

Farmland bird species utilize a variety of weed seed and invertebrate groups. Using the counts data for weed seeds and invertebrates in suction samples in the FSEs we compared the amount of ‘bird food’ on the conventionally herbicide-managed with GMHT-managed halves. We found that weed seed resource availability was reduced on the GMHT halves for most farmland bird species analysed (Figure 2a; Gibbons et al. 2006). Generally, GMHT maize yielded greater weed seed resource, while GMHT beet and spring rape yielded lower weed seed bird food. For invertebrate food for nestlings, there were relatively few effects of GMHT management (Figure 2b). Rather, as for the findings of the community analyses, invertebrates bird food was relatively insensitive to a change in herbicide mangement. Amalgamating invertebrate groups by their trophic function also allowed us to conclude that GMHT cropping and management would not affect the trophic functioning of farmland were it allowed to go ahead (Figue 2c). This work demonstrates how we can link the impacts of management to different components of a functioning system and also extrapolate to groups, such as birds, for which it is often difficult to obtain empirical data.

To evaluate how changes in management affected trophic functioning, we developed a simple linear regression methodology (Hawes et al. 2003, Bohan et al. 2007). We found evidence for strong relationships between trophic groupings that explained a considerable amount of between-crop and within-field variation, particularly where natural enemies are included in the trophic interaction (Figure 3). This work suggested that invertebrate abundance is governed both by bottom-up effects from the crop, and top-down effects from natural enemies. However, it most clearly showed why, despite significant changes in weed abundance with GMHT cropping, there were few changes in the invertebrate groups.

The outcome of these analyses formed the basis of a BBSRC funded project under Defra’s Sustainable Arable LINK programme, in collaboration between the EBDG at Rothamsted Research, Cathy Hawes at SCRI and Alan Raybould and Pernille Thorbek at Syngenta. The aim of this ongoing project is to use a combination of trophic and functional approaches to explain the effects of management on the farmland ecosystem, which we define as the crop, weed and invertebrate constituents of farmland (Figure 4).

10

a)

b)

c)

Figure 2. a) Total seed rain, per m2, of weeds important in the diets of 7 species of farmland birds in beet, maize, spring and winter oil seed rape. Values given are the treatment ratio, R (see Gibbons et al., 2006). b) Total count, per m2, of invertebrates important in the diets of 7 species of farmland birds in beet, maize, spring and winter oil seed rape. Values given are the treatment ratio, R. c) Total count, per m2, of invertebrates trophic groups in beet, maize, spring and winter oil seed rape. Values given are the treatment ratio, R (Hawes et al., 2003; Bohan et al., 2005).

11

Figure 3. Plots of a sub-set of best fitted linear regressions of trophic consumer on resource for year totals. In each case the dependent variable in the regression was chosen as the group to which energy flowed, the consumer, and the independent variable as the group from which energy flowed, the resource. The best models, shown for each relationship, were a single line, or two parallel lines, or two lines with different regression coefficients. Herbicide management was non-significant in all cases and consequently only the spring-sown oilseed rape (solid line) and winter-sown oilseed rape (dashed line) relationships are shown. The graphs are plotted only over the ranges of counts/biomass/cover observed for that relationship, with no extrapolation. The axes run, in all cases of counts and biomass, from -1 ≤ log(x) ≤ 5 along the resource axis and -1 ≤ log(y) ≤ 4 along the consumer axis. For crop cover the axes run from -1 ≤ log(x) ≤ 2 and -1 ≤ log(y) ≤ 4.

12

specialist predator

generalist predator

sap-sucker

leaf chewer

detritivores

seed eater

sap

leavesseeds

detritus

mortality

predation

herbivory

Relevant trophic traitsRepresentative trophic species Data from field & lab experiments

IBM simulations

PredictionsManagement applications

Figure 4. Schematic for a putative trophic network within a ecosystem model. Each individual in the ecosystem has its own behaviour, determined by rules that depend upon the trophic group it belongs to.

The main objective, using an Individual Based Model, is to investigate the population

dynamics of the food web (flows of energy between producer and consumer levels) and simulate effects of management on trophic groups to identify the balance between crop growth and functional diversity, and gain insights into key functional processes. We will be able to predict the influence and risks of management on plant and invertebrate communities, examine the trade-off between input costs, production yields and biodiversity to predict which management combinations are most likely to achieve maximum yields with minimum loss in diversity. We intend that this will be of use in informing government policy, product development and immediate management.

We conclude that there are now techniques for the:

1. Detection of management at various scales (species/taxa; community; functional) 2. Scaling from measured to non-measured data (e.g. Birds) 3. Understanding relationships of trophic structuring, energy flow and ecosystem

function

13

Acknowledgements The Farm Scale Evaluations were funded by Defra and the Scottish Executive. Rothamsted Research receives grant-aided support from the BBSRC. References Bohan, D.A., Boffey, C.W.H., Brooks, D.R., Clark, S.J., Dewar, A.M., Firbank, L.G., Haughton,

A.J., Hawes, C., Heard, M.S., May, M.J., Osborne, J.L., Perry, J.N., Rothery, P., Roy, D.B., Scott, R.J., Squire, G.R., Woiwod, I.P. & Champion, G.T. 2005: Effects on weed and invertebrate abundance and diversity of herbicide management in genetically modified herbicide-tolerant winter-sown oilseed rape. Proceedings of the Royal Society B-Biological Sciences 272 (1562): 463-474.

Bohan, D.A., Hawes, C., Haughton, A.J., Denholm, I., Champion, G.T., Perry, J.N. & Clark, S.J. 2007: Statistical models to evaluate invertebrate-plant trophic interactions in arable systems. Bulletin of Entomological Research 97: 265-280.

Gibbons, D.W., Bohan, D.A., Rothery, P., Stuart, R.C., Haughton, A.J., Scott, R.J., Wilson, J.D., Perry, J.N., Clark, S.J., Dawson, R.J.G. & Firbank, L.G. 2006: Weed seed resources for birds in fields with contrasting conventional and genetically modified herbicide-tolerant crops. Proceedings of the Royal Society B-Biological Sciences 273: 1921-1928.

Hawes, C., Haughton, A.J., Osborne, J.L., Roy, D.B., Clark, S.J., Perry, J.N., Rothery, P., Bohan, D.A., Brooks, D.R., Champion, G.T., Dewar, A.M., Heard, M.S., Woiwod, I.P., Daniels, R.E., Young, M.W., Parish, A.M., Scott, R.J., Firbank, L.G. & Squire, G.R. 2003: Responses of plants and invertebrate trophic groups to contrasting herbicide regimes in the Farm Scale Evaluations of genetically modified herbicide-tolerant crops. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 358(1439): 1899-1913.

Smith, V., Bohan, D.A., Clark, S.J., Haughton, A.J., Bell., J.R. & Heard, M.S. (submitted) Weed and invertebrate compositions in arable farmland.

14

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 15-22 Integrating insect-resistant GM crops in pest management systems Steven E. Naranjo USDA-ARS, Arid-Land Agricultural Research Center, 21881 North Cardon Lane, Maricopa, Arizona, USA 85226 (E-mail: Steve.Naranjo@ars.usda.gov) Abstract: In 2006, 102 million hectares of GM crops were produced globally. GM cotton and maize with insect resistance were grown on 12.1 and 20.1 million hectares in 9 and 13 countries, respectively with these crops collectively representing about 32% of all GM crops grown in 2006. These insect resistant GM crops produce various Cry toxins from Bacillus thuringiensis (Bt) and provide for highly selective and effective control of lepidopteran and coleopteran pests, primarily bollworms, borers and rootworms, which are the most damaging pests of cotton and maize worldwide. It is estimated that between 1996 and 2005 the deployment of Bt cotton and maize has reduced the volume of insecticide active ingredient used for pest control by 94.5 and 7.0 million kg and increased farm income through reduced costs and improved yields by US$7.51 and 2.37 billion, respectively. For cotton and maize pests susceptible to Bt toxins, these GM crops are an extremely successful form of host plant resistance, one of many pest management tactics that can be integrated in pest management systems. Reductions in insecticide use through adoption of Bt crops have broadened opportunities for biological control of all cotton and maize pests but most other pest management tactics have remained largely unchanged or modified only slightly in Bt crops. Many studies have clearly demonstrated enhanced natural enemy abundance in Bt crops compared with conventional crops subject to broad-spectrum chemical insecticides. A few studies also have focused on understanding the functional contribution of this natural enemy conservation. In both systems, several non-target pests have become more problematic in Bt crop fields in some countries largely due to reductions in insecticide use for target pests. Changes in IPM practice, enhanced biological control and the emergence of nontarget pests are further illustrated by examples from the Bt cotton system. Key words: IPM, Bt cotton, Bt maize, biological control, natural enemy conservation, nontarget pests, nontarget effects, decision aids

Introduction The adoption and use of genetically-modified (GM) crops continues to grow rapidly worldwide. As of 2006, 22 countries were producing GM crops on a total of 102 million hectares (James, 2006). On a percentage basis, herbicide tolerant soybeans represent about 57% of GM crop production worldwide followed by GM maize (25%), GM cotton (13%) and herbicide tolerant canola (5%) and alfalfa (<1%). Cotton and maize genetically modified to produce the selective toxin proteins of Bacillus thuringiensis (Cry proteins) have been grown commercially since 1996. Bt cotton is grown in nine countries and Bt maize in 13 countries; only Argentina, South Africa and the USA commercially grow both Bt cotton and maize. Seven countries within the European Union are producing GM crops, albeit at low levels. In 2006, Romania grew about 145,000 ha of GM soybean, Spain grew roughly 60,000 of Bt maize and France, Germany, Portugal, Czech Republic and Slovakia each grew less than 5000 ha of Bt maize (James, 2006). The adoption rate of Bt cotton in India has been unprecedented. Production there grew from about 50,000 ha in 2002, the initial year of approval, to 3.8 million ha in 2006, a 7500% increase in 4 years. India now grows more Bt cotton than any other country in the world.

15

16

The potential economic and environmental benefits of insect resistant GM crops are great. Based on farm-scale studies, Brookes and Barfoot (2006) estimate that over the 10 year period from 1996-2005, Bt cotton and maize production has increased farm income globally by US$7.51 and 2.37 billion, respectively. Not surprisingly, the USA is a large beneficiary of Bt cotton and maize because of the large production of these crops, but farm communities in India and especially China greatly benefit from production of Bt cotton as well. Also not surprising given the world distribution of Bt cotton, nearly 80% ($1.38 billion) of the income benefits were garnered by farmers in developing nations in 2005. Environmental gains are impressive as well. On a global scale, Brookes and Barfoot (2006) estimated that the total volume of insecticide active ingredient use declined by 94.5 and 7 million kg from production of Bt cotton and maize, respectively, during the period 1996-2005. Moreover, if aspects of the environmental toxicity of the insecticides used are accounted for through the environmental impact quotient (EIQ, Kovach et al., 1992), there was a 24.3 and 4.6% reduction in this quotient over the 10 year period for Bt cotton and maize, respectively. The overall lower gains in farm income and insecticide use reduction for Bt maize likely result from the fact that borers are seldom treated with insecticides due to low efficacy and Bt maize targeting corn rootworms, a pest demanding most of the insecticide use in maize, has only been commercially available since 2003. IPM in Bt crops

Target pests Modern pest control is guided by the principles of integrated pest management (IPM). The use of GM crops, which have biological activity against select insect pests, qualifies as one of the many tactics that can be integrated into IPM strategies. Within this context, Bt crops are virulent and selective forms of host plant resistance effective against the most significant pests of cotton and maize. The primary targets of Bt cottons are species of the bollworm/budworm complex (Heliothis and Helicoverpa), the pink bollworm, Pectinophora gossypiella, and various spiny and spotted bollworms (Earias spp.). The main targets of Bt maize are the borers Ostrinia nubilalis and Sesamia nonagrioides, the earworms, Helicoverpa zea and the rootworms, Diabrotica spp. These Bt crops may have activity against other lepidopteran and coleopteran pests. For example, Bt cottons also have good activity against pests such as leafworms, leaf perforators, semiloopers and other bollworms (Benedict & Ring, 2004). The spectrum and efficacy of pest lepidopteran control in cotton has been enhanced even further with dual-gene Bt cottons that produce two different Cry toxins (e.g. Bollgard II, WideStrike) (Fitt & Wilson, 2000; Adamczyk et al., 2001). These dual-gene Bt cottons are being utilized more widely in several countries such as Australia and the USA. In addition, single or double Bt gene events are now frequently stacked with genes conferring tolerance to certain herbicides.

The remainder of this paper will discuss various IPM related issues exemplified in the Bt cotton system. Nontarget pests A relatively large number of pest species that are not susceptible to the Bt toxins expressed in transgenic cottons affect production worldwide. In general, many of these species exhibit the same pest status and continue to be managed identically in Bt and conventional cropping systems. However, the use of Bt cottons have led to indirect effects on some of these nontarget pest species in some productions systems. In Australia, the reduced use of insecticides for bollworms has allowed some pests to become more prominent, the most

17

significant of which is the green mirid, Creontiades dilutus, (Lei et al., 2003), which is now sprayed as many as three times per season (Doyle et al., 2006) with broad spectrum insecticides (Khan et al., 2006). The use of these insecticides have in turn disrupted natural enemies leading to increased problems with spider mites, cotton aphids, and whiteflies (Wilson et al., 1998; Farrell et al., 2006). In northern China a complex of mirid plant bugs have recently become key insect pests in Bt cotton fields (Wu et al., 2002) and leafhoppers, cotton aphids, and spider mites have become problematic in some areas as well (Deng et al., 2003; Men et al., 2005). Likewise, mirid plant bugs and stinkbugs have risen in pest status since the adoption of Bt cottons in the mid-south and southeastern production areas of the USA (Williams, 2006). In India, the reduction in insecticide sprays, especially during both the flowering and boll formation phases, has been associated with resurgence of some minor pests such as mealy bugs, thrips and leafhoppers (Sharma et al., 2005).

The reduction of insecticide use in Bt cotton is the likely factor explaining resurgence in some nontarget pests, but other factors may be involved as well. Reduced competition from target species may enable nontarget pest populations to thrive. Negative effects of Bt cotton on natural enemy populations might also lead to enhanced nontarget pest problems. However, the bulk of evidence to date suggest that Bt crops are highly selective and that negative effects, if any, are relatively minor in magnitude (Naranjo et al., 2005; Romeis et al., 2006; Marvier et al., 2007). Thus, it is more likely that problematic nontarget pests are not under good biological control even in conventional systems. Decision aids Pest monitoring and the use of economic thresholds are fundamental components of most IPM programs. Ironically, a grower’s decision to employ Bt cotton for caterpillar control is made well in advance of actual pest population assessment and so its use is most often associated with production areas where caterpillar pests are a consistent problem year after year. With the exception of greater vigilance and awareness of plant bugs and other pests that been observed to be problematic in some Bt cotton systems, identical decision-making protocols generally apply to nontarget pests in both Bt and conventional production fields. Bt cottons producing Cry1Ac toxins are not completely effective against Helicoverpa spp., thus, growers must continue to monitor populations of these pests. In most cases there have been slight modifications to sampling and thresholds protocols for the major target pests in Bt cotton fields. For example, in the USA sampling concentrates on older (2-3 days old) larvae and use of a lower threshold level than conventional cotton. Focus on slightly older larvae helps to identify populations not being effectively controlled by the Bt toxins but still of a size amenable to control with available insecticides (Farm Press, 2006). The lower threshold reflects the fact that larger larvae, which are the focus of sampling, are capable of greater damage if left untreated. Similar strategies are employed for Helicoverpa in Australia and China (Wu & Guo, 2005; Farrell et al., 2006). Biological control As a single IPM tactic, Bt cotton provides obvious opportunities for improved caterpillar control, but the interactions and synergies from a single tactic such as host plant resistance may provide benefits beyond its narrow range of direct control on a specific group of pest species. Thus, although Bt cotton directly controls only lepidopteran pests, the associated reduction in insecticide use for these pests may facilitate or enhance the effectiveness of other tactics, most notably, biological control which may directly contribute to control of other pests in the system. Cotton supports a large and diverse assemblage of arthropod natural enemies (Whitcomb & Bell, 1964; Bishop & Blood, 1977; Zhao, 1984; Romeis & Shanower,

18

1996; Dippenaar-Schoeman et al., 1999) and many studies have demonstrated that natural enemies play a critical role in suppression of cotton pests (e.g., Eveleens et al., 1973; Abdelrahmin & Munir, 1989; Trichilo & Wilson, 1993; Wilson et al., 1998; Sharma et al., 2007). Reduced insecticide use due to adoption of Bt-transgenic cotton, the growing availability and use of selective insecticides, and improvements in other pest management tactics, have created significant opportunities for biological control in the cotton system (Figure 1)

1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Inse

ctic

ide

appl

icat

ions

/acr

e

2

4

6

8

Figure 1. Insecticides use patterns in the USA 1986-2006 relative to all pests. Arrow denotes the beginning of Bt cotton production. Compiled from data of the National Cotton Council (http://www.cotton.org/tech/pest/index.cfm).

Conservation biological control, which relies on the preservation, manipulation and/or enhancement of existing natural enemies in the system, is the most widely practiced and successful form of biological pest control in cotton worldwide. Classical and augmentative biological control have played much smaller roles in cotton pest control (King et al., 1996), although some augmentation biological control is used in China and India (Sharma, 2005; Wu & Guo, 2005). Conservation biological control is largely achieved with relatively little or no overt action by growers other than consideration of the types of insecticides used, when necessary, to achieve pest control. Use of Bt cotton along with other changes in pest management practices have led to generally higher populations of natural enemies in cotton systems in many parts of the world (e.g., Wilson et al., 2004; Wu & Guo, 2005; Sharma & Ortiz, 2000; Naranjo et al., 2004).

The tangible benefits of improved conservation of natural enemy populations in Bt cotton have been demonstrated in several systems. In northern China, Wu and Guo (2003) have shown that cotton aphids that are resistant to various insecticides used to control bollworms in cotton are effectively suppressed by natural enemies in Bt cotton fields where such sprays are unnecessary. In contrast, insecticides used to control bollworms in non-Bt cotton field disrupt natural enemies leading to outbreaks of aphids. In the western USA, whiteflies (Bemisia tabaci) can be suppressed long term in cotton fields with a single initial application of selective insecticides in comparison with fields using broad spectrum insecticides where repeated applications are necessary. Bt cotton underpins this successful whitefly management strategy by reducing or eliminating sprays needed to control caterpillar pests (Naranjo, 2001; Ellsworth & Martinez-Carrillo, 2001). In fact, the widespread adoption of Bt cotton in the western USA has had a large negative effect on regional populations of pink bollworm, the

19

main Bt cotton target in this area (Carrière et al., 2003) and has dramatically reduced insecticide use for all pests in cotton (Ellsworth and Jones, 2001). This reduction has undoubtedly been fueled in part by the biological control imposed by abundant natural enemy populations facilitated by use of Bt cotton.

As to the effect of Bt cotton itself on natural enemies, numerous studies have failed to find meaningful impacts (reviewed by Romeis et al., 2006) and recent meta-analysis by Marvier et al. (2007) based on nontarget field studies conducted worldwide generally support this conclusion, and further highlight the much more detrimental effect of insecticide use. Few studies have examined the relative biological control capacities of Bt and non-Bt cotton fields. However, in no case to date has biological control capacity been reduced in Bt cotton fields compared to non-Bt fields. For example, studies in Arizona cotton revealed that the abundance of lygus bugs and whiteflies, two primary pests of cotton in the region, and predator:prey ratios for these pests were identical in Bt and non-Bt cotton fields that were not subject to any insecticides (Naranjo, 2005a,b). Further functional studies in this system showed that rates of mortality imposed by natural enemies (primarily predators) on both target and nontarget pests were again identical in Bt and non-Bt cotton fields (Naranjo, 2005b). Conclusions The rate and scale of adoption of insect-resistant GM crops is unprecedented relative to other advances in production. For Bt cotton, the rapid adoption by a wide cross-section of growers, large and small, has been largely driven by the significant economic benefits of the technology in reducing production costs while improving yield and quality. The technology also has had dramatic positive impacts on the environment globally through the reduction in insecticide usage in a system that has historically been associated with insecticide over-reliance and misuse. Bt cotton represents only one of a myriad of tactics that can be integrated into efficient and effective pest management strategies; however, its contribution via reduction in insecticide usage has the potential to cascade through the system and enhance other integrated pest management tactics, most notably biological control. Numerous non-target studies have definitively shown the selective nature of this pest control technology and indicate that enhanced biological control should be possible. Evidence from several regions have shown the key role of biological control in managing nontarget pests in Bt cotton. In some systems the use of Bt cotton has been associated with increased pressure from other primary or secondary pests not susceptible to Bt toxins that may have been previously suppressed by insecticide use for target pest species. In these instances it is likely that effective biological control was absent initially and the selective action of Bt has not improved the situation. The adoption of Bt cotton has had only minor effects on other pest management practices such as sampling and use of economic thresholds. The deployment of Bt crops has the potential to have suppressive areawide effects on pest populations leading to reduced risk and greater predictability for growers, and improved environmental stewardship.

References Abdelrahman, A.A. & Munir, B. 1989: Sudanese experience in integrated pest management of

cotton. Insect Science and Applications 10: 787-794. Adamczyk, J.J., Jr., Adams, L.C. & Hardee, D.D. 2001: Field efficacy and seasonal

expression profiles for terminal leaves or single and double Bacillus thuringiensis toxin cotton genotypes. Journal of Economic Entomology 94: 1589-1593.

20

Benedict, J.H. & Ring, D.R. 2004: Transgenic crops expressing Bt proteins: current status, challenges and outlook. In: Transgenic Crop Protection: Concepts and Strategies, eds. O. Koul and G.S. Dhaliwal, Science Publishers, Inc., Enfield, NH, USA, pp. 15-84.

Bishop, A.L. & Blood, P.R.B. 1977: A record of beneficial arthropods in South East Queensland cotton. PANS 23: 384-386.

Brookes, G. & Barfoot, P. 2006: Global impact of biotech crops: Socio-economic and environmental effects in the first ten years of commercial use. AgBioForum 9: 139-151.

Carrière, Y., Ellers-Kirk C., Sisterson M., Antilla L., Whitlow M., Dennehy T. J.& Tabashnik B. E. 2003: Long-term regional suppression of pink bollworm by Bacillus thuringiensis cotton. Proceedings of the National Academy of Sciences 100: 1519–1523.

Deng, S.D., Xu, J., Zhang, Q.W., Zhou, S.W. & Xu, G.J. 2003: Effect of transgenic Bacillus thuringiensis cotton on population dynamics of the non-target pests and natural enemies. Acta Entomologica Sinica 46: 1-5.

Dippennaar-Schoeman, A.S., van den Berg, A.M. & van den Berg, A. 1999: Spiders in South African cotton fields: species diversity and abundance (Arachnida: Araneae). African Plant Protection 5: 93-103.

Doyle, B., Reeve, I. & Coleman, M. 2006: The CCA 2005 Bollgard comparison report: A survey of cotton growers’ and consultants’ experience with Bollgard in the 2004-2005 season.. The Cotton Research and Development Corporation and The Cotton Catchment Community Cooperative Research Centre, Institute for Rural Futures, Armidale, New South Wales, Australia.

Ellsworth, P.C. & Jones, J.S. 2001: Cotton IPM in Arizona: a decade of research, implementation & education. In: Proceedings Beltwide Cotton Conferences, National Cotton Council, Memphis, TN, pp. 1088-1096.

Ellsworth, P.C. & Martinez-Carrillo, J.L. 2001: IPM for Bemisia tabaci: a case study from North America. Crop Protection 20: 853-869.

Eveleens, K.G., van den Bosch, R. & Ehler, L.E. 1973: Secondary outbreak induction of beet armyworm by experimental insecticide application in cotton in California. Environmental Entomology 2: 497-503.

Farm Press. 2006. Delta Agricultural Digest 2006: Farm Press, Clarksdale, Mississippi, USA Farrell, T., Mensah, R., Sequeira, R., Wilson, L. & Dillon, M. 2006: Key insect and mite pests

of Australian Cotton. In: Cotton Pest Management Guide 2006-07, ed. T. Farrell, New South Wales Department of Primary Industries, pp. 1-17.

Fitt, G.P. & Wilson, L.J. 2000: Genetic engineering in IPM: Bt cotton. In: Emerging Technologies for Integrated Pest Management: Concepts, Research and Implementation, eds. G.G. Kennedy and T.B. Sutton, APS Press, St. Paul, Minnesota, pp. 108-125.

James, C. 2006: Global Status of Commercialized Biotech/GM Crops: 2006. ISAAA Briefs, No. 35, ISAAA, Ithaca, NY, USA.

Khan, M., Quade, A. & Murray, D. 2006: Mirid management - effect of salt rate when mixed with reduced rates of chemical. In: Proceedings of the 13th Australian Cotton Conference, Australian Cotton Growers Research Association, Narrabri, New South Wales, Australia, pp. 537-542.

King, E.G., Coleman, R.J., Morales-Ramos, J.A., Summy, K.R., Bell, M.R. & Snodgras, G.L. 1996: Biological control. In: Cotton Insects and Mites: Characterization and Management, eds. E.G. King, J.R. Phillips and R.J. Coleman, The Cotton Foundation Publisher, Memphis, Tennessee, pp. 511-538.

21

Kovach, J., Petzoldt, C., Degni, J. & Tette, J. 1992: A method to measure the environmental impact of pesticides. New York Food and Life Sciences Bulletin, New York State Agricultural Experiment Station, Cornell University, Geneva, NY (http://www.nysipm.cornell.edu/publications/eiq/default.asp, accessed 30 July 2007)

Lei, T., Khan, M. & Wilson, L. 2003: Boll damage by sucking pests: An emerging threat, but what do we know about it? In: World Cotton Research Conference III: Cotton for the New Millennium, ed. A. Swanepoel, Agricultural Research Council - Institute for Industrial Crops, Cape Town, South Africa, pp. 1337-1344.

Marvier, M., McCreedy, C., Regetz, J. & Kareiva, P. 2007: A meta-analysis of effects of Bt cotton and maize on nontarget invertebrates. Science 316: 1475-1477.

Men, X., Ge, F., Edwards, C.A. & Yardim, E.N. 2005: The influence of pesticide applications on Helicoverpa armigera (Hübner) and sucking pests in transgenic Bt cotton and non-transgenic cotton in China. Crop Protection 24: 319-324.

Naranjo, S.E. 2001: Conservation and evaluation of natural enemies in IPM systems for Bemisia tabaci. Crop Protection 20: 835-852.

Naranjo, S.E. 2005a: Long-term assessment of the effects of transgenic Bt cotton on the abundance of nontarget arthropod natural enemies. Environmental Entomology 34: 1193-1210.

Naranjo, S.E. 2005b: Long-term assessment of the effects of transgenic Bt cotton on the function of the natural enemy community. Environmental Entomology 34: 1211-1223.

Naranjo, S.E., Ellsworth, P.C. & Hagler, J.R. 2004: Conservation of natural enemies in cotton: role of insect growth regulators in management of Bemisia tabaci. Biological Control 30: 52-72.

Naranjo, S.E., Head, G. & Dively, G.P. 2005: Field Studies Assessing Arthropod Nontarget Effects in Bt Transgenic Crops: Introduction. Environmental Entomology 34: 1178-1180.

Romeis, J., Meissle, M. & Bigler, F. 2006: Transgenic crops expressing Bacillus thuringiensis toxins and biological control. Nature Biotechnology 24: 63-71.

Romeis, J. & Shanower, T.G. 1996: Arthropod natural enemies of Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae) in India. Biocontrol Science and Technology 6: 481-508.

Sharma, H.C., (Ed.) 2005: Heliothis/Helicoverpa Management: Emerging Trends and Strategies for Future Research. Oxford and IBH Publishers, New Delhi, India

Sharma, H.C. & Ortiz, R. 2000: Transgenics, pest management and the environment. Current Science 79: 421-437.

Sharma, H.C., Arora, R. & Pampapathy, G. 2007: Influence of transgenic cottons with Bacillus thuringiensis cry1Ac gene on the natural enemies of Helicoverpa armigera. BioControl 52: 469-489.

Sharma, O.P., Bambawale, O.M., Dhandapani, A., Tanwar, R.K., Bhosle, B.B., Lavekar, R.C. & Rathod, K.S. 2005: Assessment of severity of important diseases of rainfed Bt transgenic cotton in Southern Maharashtra. Indian Phytopathology 58: 483-485.

Trichilo, P.J. & Wilson, L.T. 1993: An ecosystem analysis of spider mite outbreaks: Physiological stimulation or natural enemy suppression. Experimental and Applied Acarology 17: 291-314.

Whitcomb, W.H. & Bell, K. 1964: Predaceous insects, spiders and mites of Arkansas cotton fields. Arkansas Agricultural Experiment Station Bulletin 690: 1-84.

Williams, M.R. 2006: Cotton insect losses 2005. In: Proceedings of the Beltwide Cotton Conferences, National Cotton Council, Memphis, TN, pp. 1151-1204.

22

Wilson, L.J., Bauer, L.R. & Lally, D.A. 1998: Effect of early season insecticide use on predators and outbreaks of spider mites (Acari: Tetranychidae) in cotton. Bulletin of Entomological Research 88: 477-488.

Wilson, L.J., Mensah, R.K. & Fitt, G.P. 2004: Implementing integrated pest management in Australian cotton. In: Novel Approaches to Insect Pest Management in Field and Protected Crops, eds. A.R. Horowitz and I. Ishaaya, Springer-Verlag, Berlin, pp. 97-118.

Wu, K.M. & Guo, Y.Y. 2003: Influences of Bacillus thuringiensis Berliner cotton planting on population dynamics of the cotton aphid, Aphis gossypii Glover, in northern China. Environmental Entomology 32: 312-318.

Wu, K.M. & Guo, Y.Y. 2005: The evolution of cotton pest management practices in China. Annual Review of Entomology 50: 31-52.

Wu, K., Li, W., Feng, H. & Guo, Y. 2002: Seasonal abundance of the mirids, Lygus lucorum and Adelphocoris spp. (Hemiptera: Miridae) on Bt cotton in northern China. Crop Protection 21: 997-1002.

Zhao, J.Z. 1984: Species distribution and population fluctuation of predacious spiders in cotton fields in China. Natural Enemies of Insects 6: 1-12.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 23-29 Impact of glyphosate use on arthropods in transgenic herbicide-tolerant maize; preliminary results from studies in Spain Ramon Albajes, Matilde Eizaguirre, Daniel Casado, Meritxell Pérez, Carmen López, Belén Lumbierres, Xavier Pons Universitat de Lleida, Centre UdL-IRTA, Rovira Roure 191, 25198 Lleida, Spain (E-mail: ramon.albajes@irta.es) Abstract: In 2006, a 4-year farm-scale study on potential impacts of glyphosate-tolerant maize on maize arthropods was initiated in Spain with the sponsorship of the Spanish Ministry of Environment. A four-block complete randomised design with 0.5 ha elementary plots and two treatments (treated with glyphosate twice and untreated) was used. Plots were sampled 7 times during the season with visual inspections, pitfall traps and yellow sticky traps. As expected, weeds in untreated plots were more abundant and different in species composition in comparison with treated plots. On-plant countings revealed that the prevalent herbivores—leafhoppers and aphids— and predatory groups -Orius spp., spiders and trombidids- were significantly more abundant in treated plots whereas the remaining groups of predators were not significantly affected. Leafhoppers were also more abundant on the yellow sticky traps located in treated plots but differences were lower than in on-plant countings. Yellow sticky traps recorded more mymarids (as observed in visual countings) in treated plots but fewer braconids and ichneumonids; chalcidids showed no differences. In contrast with the results of visual samplings, there were significantly more Orius spp., staphylinids and thrips (both predatory and herbivore ones) on yellow traps placed in untreated than in treated plots. Among soil dwelling predators, spiders and carabids were caught significantly more in pitfall traps of untreated plots, whereas rove beetle and earwig numbers showed no difference between the two types of plot. Catches of elongate collembola were higher in untreated plots but the differences were not significant. Key words: GMHT crop, corn, predator, parasitoid, herbivore. Introduction Tolerance to herbicides is the most common transgenic trait introduced into commercial genetically modified crops worldwide, reaching 83 million hectares—alone or stacked with Bt traits—in 2006 (James, 2006). This acreage represented 81% of the total global area cultivated with transgenic crops that year. At the moment GMHT crops are not allowed in Europe for cultivation.

Herbicide-tolerant (HT) crops enable the use of broad-spectrum herbicides that could lead in the short or long term to substantial changes in the abundance, composition and phenology of weed flora and in subsequent trophic levels in maize fields, especially in areas with continuous maize cultivation. However, unlike transgenic Bt crops, in which the expression of the transgenic trait and its repercussions on arthropods cannot be altered to a great extent by the grower’s cultural practices, the impact of GMHT crops on arthropods (and other organisms) greatly depends on how herbicides are managed. Whereas a fairly high number of studies have been made on impacts of Bt maize on arthropods, far fewer have been devoted to HT maize. Papers published by Brooks et al. (2003) and Haughton et al. (2003) have focused on questions of arthropod diversity, but potential consequences for biological control functions have hardly been discussed.

23

24

To investigate the short- and long-term effects of repeated use of glyphosate on maize weed flora and arthropod fauna as a consequence of the deployment of GMHT maize in Spain, the National Institute of Agriculture and Agrofood Technology (INIA) and the University of Lleida (UdL) are developing a 4-year study. The first year (2006) was devoted to acquiring some baseline data on the arthropod and functional groups (herbivores, predators, parasitoids and decomposers) that may be more sensitive to changes in herbicide management practices. The most significant results are here reported. Material and Methods Experimental design and treatments A complete random block design with one treated (T) and one untreated (NT) plot and 4 replications was used. The elementary plots were 0.5 ha in size. Treatment consisted of two applications of glyphosate at V4 and V8 maize growth stages at a rate of 3 l/ha (1.08 g/ha of a.i.). The remaining cultural practices were the same as those on the untreated plot and corresponded to those common in the region. Seed was dressed (the only insecticide treatment applied) with imidacloprid. The whole experimental field was sown with the same variety based on the transformation event NK603. Sampling Three techniques were used to estimate arthropod densities or activities. For visual countings, 25 plants per plot were inspected. Three pitfall traps (a glass jar of 9 cm ∅ and 17 cm depth half-filled with water, 20% ethylene-glycol and detergent) were installed in each plot, regularly distributed along the plot length, and left active for 5-6 days. Three yellow stick traps (21 x 31 cm, only one sticky side, Serbios®, Italy) per plot were put on a stake at canopy height (until V12) or at ear level (from V15 onwards) and left active for 3-4 days. Samples were taken 7 times with each of the techniques at the following maize growth stages: V6-7, V8-10, V12-14, V14-15, R1, R3 and R5 [nomenclature of Ritchie et al. (1992)]. Our colleagues in the weed science group (INIA, Madrid) sampled each plot three times to estimate weed density and composition. For this, weeds inside 40 circles (∅ 31 cm) per plot were counted and identified. Statistical analysis The number of arthropods were transformed by SQRT(x+0.5) for normalisation purposes. A two-way (T vs. NT and sampling date) ANOVA with repeated measures was used to analyse the influence of the treatment on arthropod numbers and a LSD test was used to separate means. Results and Discussion Weed flora abundance and composition The most abundant weeds were: Echinochloa spp., Setaria spp. (both genera recorded together, ES), Portulaca oleracea L. (Po), Veronica persica Poiret (Vp), Amaranthus retroflexus L. (Ar), Medicago sativa L. (Ms), Lolium spp. (Ls), Capsella bursa-pastoris L. (Cbp), Cyperus rotundus L. (Cr), Rumex spp.(Rs) and Abutilon theophrasti Medik (At). As expected, weeds were much more abundant in NT plots (97.2 plants/m2) than in the plots treated twice with glyphosate (5.5 plants/m2) (all weed data were provided by García Baudín et al., personal communication). In addition, weed composition differed between the two types of plot: whereas in NT plots ES was predominant (65%) and Vp was the second (10%), in T plots ES only represented 13% of the total weed community, preceded by Ms (36%) and Po (31%). It should

25

be noted that alfalfa (Ms) was the previous crop in the experimental field, possibly explaining its high abundance in T plots. Visual sampling While in treated plots there were a higher proportion of herbivores (76%) than in untreated plots (64%), the opposite tendency was observed for predators. Statistical analyses of major arthropod groups recorded on plants are shown in Table 1.

Among the three prevalent herbivore groups, aphids (Table 1) and particularly leafhoppers [mostly Zyginidia scutellaris (Herrich-Schäffer)] were more abundant (P<0.05) on plants of T plots, whereas thrips did not show significant differences. Differences in abundance of the two herbivore groups were mainly found on three of the seven sampling dates (Figure 1). Among the prevalent predatory groups recorded on maize plants, Orius spp., spiders and trombidids (about 50% of the total predators recorded) were more abundant in T than in NT plots (Table 1), differences found in the second part of the season for Orius and spiders (Figure 1) and during most of the season for trombidids. Table 1. ANOVAs of the influence of the glyphosate treatment (T) and sampling date (S) on the abundance (mean of individuals/plant±S.E.) of different groups of arthropods recorded on maize plants. Means within a row with no letters are not significantly different (P<0.05).

Treatment (T) Group treated untreated Pa Sampling (S), Pb TxS, Pc

HERBIVORES Total aphids 2.41±0.45a 1.53±0.43b 0.05 0.001 0.19 Herbivorous thrips 3.55±0.80 3.30±0.71 0.81 <0.0001 0.19 Leafhoppers 24.92±5.78a 8.11±1.98b (0.01) <0.0001 0.0001 PREDATORS Total predators 8.64±1.15 7.37±1.14 0.27 <0.0001 0.23 Orius spp. 1.99±0.29 1.81±0.27 0.23 <0.0001 0.02 Total Nabis <0.01±<0.01b 0.25±0,10a 0.03 0.09 0.08 Mirids 1.02±0.23 0.97±0.19 0.81 <0.0001 0.25 Coccinellids 1.42±0.58 0.61±0.25 0.27 0.03 0.74 Carabid adults 0.32±0.10 0.46±0.11 0.06 <0.0001 0.04 Staphylinid adults 0.65±0.10 0.65±0.12 0.84 <0.0001 0.72 Chrysopids 1.45±0.20 1.51±0.25 0.99 0.0001 0.28 Syrphids 0.63±0.23 0.95±0.36 0.16 <0.0001 0.08 Spiders 1.35±0.24a 0.83±0.12b 0.001 <0.0001 0.002 Predatory thrips 0.68±0.17 0.63±0.12 0.81 <0.0001 0.80 Trombidids 1.02±0.13a 0.62±0.10b 0.02 0.02 0.10

ad.f.= 1,3; bd.f.=6,18; cd.f.=6,18

26

Aphids

0

2

4

6

8

V6 V9 V13 V15 R1 R3 R6

Growth Stage

indi

vidu

als/

plan

tNTT

Leafhoppers

0

20

40

60

80

V6 V9 V13 V15 R1 R3 R6

Growth Stage

indi

vidu

als/

plan

t

NTT

Orius

0

1

2

3

4

5

V6 V9 V13 V15 R1 R3 R6

Growth Stage

indi

vidu

als/

plan

t

NTT

Spiders

0

1

2

3

4

5

V6 V9 V13 V15 R1 R3 R6

Growth Stagein

divi

dual

s/pl

ant

NTT

Figure 1. Mean (±SE) number/plant of main arthropod groups during the season in plots treated (T) twice with glyphosate and in untreated plots (NT) as determined by visual sampling. Pitfall traps Pitfall trap records showed a very different picture than visual samplings. The predominant predatory groups were more active in NT plots although differences were not statistically significant in the case of staphylinids and earwigs (Table 2). Table 2. ANOVAs of the influence of the glyphosate treatment (T) and sampling date (S) on the abundance (mean of individuals/trap±S.E.) of different groups of arthropods caught in pitfall traps. Means within a row with no letters are not significantly different (P<0.05).

Treatment (T) Group treated Untreated Pa Sampling (S), Pb TxS, Pc

Total carabids 1.41±0.49b 8.26±1.82a 0.006 0.009) 0.15 Total staphylinids 5.76±0.77 8.52±1.18 0.22 <0.0001 0.10 Earwigs 0.46±0.18 0.83±0.23 0.18 <0.0001 0.53 Spiders 8.53±0.91b 15.25±1.71a 0.05 <0.0001 0.02 Centipedes/millipedes 11.95±3.82 11.55±2.85 1.00 <0.0001 0.58 Collembola, globular 0 0.01±0.01 0.39 0.46 0.46 Collembola, elongate 12.94±6.55 33.17±17.68 0.51 0.06 0.91

ad.f.= 1,3; bd.f.=6,18; cd.f.=6,18

27

Differences between treatments for carabids and spiders were consistent during all or most of the season (Figure 2). The differences recorded in carabids were particularly relevant, given that they were extremely low in T plots but reached considerably higher numbers in NT plots. Elongate collembola, as representative of detritivorous insects, caught in untreated plots were 2.5 times higher in number than those caught in treated plots but the differences were not statistically significant, mainly due to the high variability of catches.

Carabids

0

5

10

15

20

25

30

V6 V9 V13 V15 R1 R3 R6

Growth Stage

indi

vidu

als/

plan

t

NTT

Spiders

0

5

10

15

20

25

30

V6 V9 V13 V15 R1 R3 R6

Growth Stagein

divi

dual

s/pl

ant

NTT

Figure 2. Mean (±SE) number of carabids and spiders (all species together) caught per pitfall trap during the season in plots treated (T) twice with glyphosate and in untreated plots (NT).

Yellow sticky traps Main herbivores recorded on sticky traps were the same as those observed in visual countings: aphids, thrips and leafhoppers. However, differences between glyphosate treated and untreated plots were not the same as those recorded on maize plants in visual countings, probably due to the capture by yellow sticky traps of flying insects coming from weed plants. This conclusion is also valid for predatory insects, such as Orius spp. and staphylinids, which showed higher catches in NT plots (Table 3).

Yellow sticky traps caught adult parasitoids mainly belonging to the families Braconidae, Ichneumonidae, Chalcididae and Mymaridae. The first two families were caught significantly more in untreated plots, chalcidids showed no differences and mymarids were caught more in traps located in treated plots. Mymarids seemed to be correlated to leafhopper abundance (as they are parasitoids of leafhopper eggs), and they could have helped to reduce the number of these homopterans after flowering, especially in treated plots. Furthermore, braconids (which include numerous aphid parasitoids) could have played a role in reducing aphid densities.

Other common insect groups caught on sticky traps, such as chloropids, showed no significant differences.

Several authors have concluded that higher diversity of weeds in crop fields leads to higher numbers of natural enemies (Altieri, 1999; Strandberg et al., 2004; Taylor et al., 2006). Our results do only partially confirm this assumption, as higher numbers of some of the most abundant predators were recorded in treated plots. Most of the work conducted for comparing weed abundance/diversity and arthropod abundance/diversity focuses on diversity questions and far less on biological control functions. Often crop plants are not sampled, or they are sampled using an inadequate technique (such as suction sampling) (e.g. Haughton et al., 2003; Strandberg et al., 2004; Taylor et al., 2006). As seen in the results here reported, visual countings on crop plants may give quite a different and more realistic picture than other sampling techniques such as yellow traps.

28

Table 3. ANOVAs of the influence of the glyphosate treatment (T) and sampling date (S) on numbers (means of individuals/trap±S.E.) of different groups of arthropods caught on yellow sticky traps. Means within a row with no letters are not significantly different (P<0.05). Raw data transformed by SQRT(x+0.5) for analyses.

Treatment Group treated untreated Pa Sampling, Pb TxS, Pb

HERBIVORES Thrips 736.1±139.0b 1545.3±342.0a 0.02 <0.0001 0.13 Leafhoppers 112.5±20.1a 85.6±13.3b 0.04 <0.0001 0.24 Aphids 8.64±2.2 11.35±2.4 0.18 <0.0001 0.11 PREDATORS Orius spp. 5.61±0.8b 7.69±1.1a 0.005 <0.0001 0.009 Staphylinids 1.40±0.25b 1.87±0.26a 0.05 <0.0001 0.67 PARASITOIDS Braconids 2.36±0.29b 4.85±0.85a 0.03 0.02 0.0008 Ichneumonids 1.89±0.44b 3.86±1.01a 0.03 0.0001 <0.001 Chalcidids 31.55±9.38 37.89±5.40 0.13 0.0001 0.49 Mymarids 120.87±45.41a 60.27±20.16b 0.01 <0.0001 0.02 OTHERS Chloropids 31.46±3.02 33.18±3.49 0.84 <0.0001 0.86 ad.f.= 1,3; bd.f.=6,18

The higher densities of aphids and leafhoppers on maize plants of treated plots could be the cause of the higher colonisation of these plots by generalist predators such as anthocorids (i.e. Orius spp.) and spiders. The mechanisms leading to enhancement of herbivore populations on maize plants of treated plots are unknown. However, different “degrees of comfort” of weed-free or weed-covered soil may be suggested. It is known that more aphids may be found on weed-free cereal plots (Way & Cammell, 1981) and that green mulches reduce aphid infestations (Döring et al., 2004). Other Homoptera do also respond in a similar way to weed-free soil (Smith, 1976) and this could also explain the higher leafhopper infestation in treated plots. In this respect, the timing of glyphosate application could be decisive for managing the abundance, composition and phenology of weed populations and therefore crop colonisation by herbivores, as has recently been hypothesised (Strandberg et al., 2004). More research is needed to determine the key processes influencing composition, abundance and dynamics of arthropod fauna in maize fields with different broad-spectrum herbicide management regimes, and particularly how repeated glyphosate use and modification of weed flora in maize affect biological control functions. This is the aim of the research planned for the coming years within this project. Acknowledgements This research is being carried out in the framework of a 4-year agreement between the public National Institute of Agriculture and Agrifood Technology (INIA) and the UdL within the agreement between the Spanish Ministry of Environment and INIA.

29

References Altieri, M.A. 1999: The ecological role of biodiversity in agroecosystems. Agric. Ecosyst.

Environ. 74: 19-31. Brooks, D.R., Bohan, D.A., Champion, G.T., Haughton, A.J., Hawes, C., Heard, M.S., Clark,

S.J., Dewar, A.M., Firbank, L.G., Perry, J.N., Rothery, P., Scott, R.J., Woiwod, I.P., Birchall, C., Skellern, M.P., Walker, J.H., Baker, P., Bell, D., Browne, E.L., Dewar, A.J.G., Fairfax, C.M., Garner, B.H., Haylock, L.A., Horne, S.L., Hulmes, S.E., Mason, N.S., Norton, L.R., Nuttall, P., Randle, Z., Rossall, M.J., Sands, R.J.N., Singer, E.J. & Walker, M.J. 2003: Invertebrate responses to the management of genetically modified herbicide-tolerant and conventional spring crops. I. Soil-surface-active invertebrates. Phil. Trans. R. Soc. B. 358: 1847-1862.

Döring, T.F, Kirchner, S.M., Kühne, S. & Saucke, H. 2004: Response of alate aphids to green targets on coloured backgrounds. Entomol. Exp. Appl. 113: 53-61.

Haughton, A.J., Champion, G.T., Hawes, C., Heard, M.S., Brooks, D.R., Bohan, D.A., Clark, S.J., Dewar, A.M., Firbank, L.G., Osborne, J.L., Perry, J.N., Rothery, P., Roy, D.B., Scott, R.J., Woiwod, I.P., Birchall, C., Skellern, M.P., Walker, J.H., Baker, P., Browne, E.L., Dewar, A.J.G., Garner, B.H., Haylock, L.A., Horne, S.L., Mason, N.S., Sands, R.J.N. & Walker, M.J. 2003: Invertebrate responses to the management of genetically modified herbicide-tolerant and conventional spring crops. II. Within-field epigeal and aerial arthropods. Phil. Trans. R. Soc. B. 358: 1863-1877.

James, C. 2006: Global status of commercialised Biotech/GM crops: 2006. ISAAA Brief No. 35. Ritchie, S.W., Hanway, J.J. & Benson, G.O. 1992: How a corn plant develops. Special Report

No. 48. Iowa State University, Ames, USA. Smith, J.G. 1976: Influence of crop background on aphids and other phytophagous insects on

Brussels sprouts. Ann. Appl. Biol. 83: 1-13. Strandberg, B., Bruus Pedersen, M. & Elmegaard, N. 2004. Weed and arthropod populations in

conventional and genetically modified herbicide tolerant fodder beet fields. Agric. Ecosyst. Environ. 105: 243-253.

Taylor, R.L., Maxwell, B.D. & Boik, R.J. 2006: Indirect effects of herbicides on bird food resources and beneficial arthropods. Agric. Ecosyst. Environ. 116: 157-164.

Way, M.J. & Cammell, M.E. 1981: Effects of weeds and weed control on invertebrate pest ecology. In: Pest, Pathogens and Vegetation. J.M. Thresh (ed.). Pitman, London, UK, pp. 443-458.

30

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 31-35 Preventing spread of Ostrinia nubilalis Hbn. by cultivation of Bt transgenic maize – First field experiments in southeastern Poland Paweł K. Bereś¹, Robert Gabarkiewicz² 1 Institute of Plant Protection, Regional Experimental Station, Langiewicza 28, 35-101Rzeszów, Poland (E-mail: p.beres@ior.poznan.pl); 2 Monsanto Poland, Domaniewska 41, 02-672 Warszawa, Poland (E-mail: robert.gabarkiewicz@monsanto.com) Abstract: During the last ten years, the area under maize cultivation has been steadily increasing from the traditional southern region to the central or even northern regions of Poland. The main pest species of maize is the European corn borer (Ostrinia nubilalis Hbn.). In the southern, warmer parts of Poland, the pest damaged 1 – 25, locally even 70% of the plants between 1957 and 1994.Recently, 60 – 80%, are damaged and some fields are completely destroyed. Recently, serious infestation level have also been reported from central regions where 10 – 15% are infested on the range boundary of O. nubilalis and there are predictions that the pest will soon cover the whole country.

The staff of the Rzeszów Experimental Station has carried out long term observations and experiments on abundance, biology, economic importance and control methods (including biological control) of maize pests, including O. nubilalis. In the 2005 and 2006 growing season the project has included observations on Bt transgenic cultivars and their parent non-Bt cultivars grown in three locations in the vicinity of Rzeszów.

Tested Bt maize cultivars demonstrated high resistance to the European corn borer. The reduction of plants damage was equal to 97.8 – 100% in comparison to the non-Bt plants. At the same time other infestation symptoms affecting yield quality and quantity (e.g. broken stem below the cob and damaged cob base by larval feeding) were also reduced for Bt plants. However, a low damage level due to the sporadic establishment of O. nubilalis larvae in maize stems of Bt cultivars was observed. When looking at economical (cost of equipment) and technical problems with chemical control of stem borers (needs for high wheel tractors and sprayers) and at the relatively unpredictable efficacy of biological control using Trichogramma spp. we believe that the cultivation of Bt maize varieties may reduce and delay the presently observed expansion of the European corn borer, which currently brings intolerable yield losses of farmers in Poland. Key words: Ostrinia nubilalis Hbn., Poland, harmfulness, transgenic maize, crop damage Introduction The European corn borer (Ostrinia nubilalis Hbn.) is the most dangerous pest of maize in Poland. Caterpillars of this butterfly pest cause damage to maize crops since the 1950s (Kania, 1962). Initially the species occurred only in southwestern Poland in the Wrocław area, where it damaged 1 to 25% of the maize plants, locally even 70% (Kania, 1961). As the land area under maize cultivation increased, the range of its occurrence broadened systematically (Lisowicz, 1995). From 1994 onwards, the caterpillars have fed on maize in the south-eastern part of the country as well (Lisowicz 2001). Over the last ten years, the range of occurrence of O. nubilalis grew quite suddenly. At present, this pest occurs already in 12 provinces: Podkarpackie, Lubelskie, Małopolskie, Świętokrzyskie, Mazowieckie, Śląskie, Opolskie, Dolnośląskie, Lubuskie, Wielkopolskie, Łódzkie and Zachodnio-Pomorskie (Bereś & Kaniuczak, 2006), causing most losses in southern Poland, where it damages 60 to 80%, and

31

32

locally even up to 100% of maize plants (Lisowicz & Tekiela, 2004). It is predicted that in the next few years the caterpillars will cause problems in maize crops all over Poland.

With the high potential of damage, it becomes necessary to control O. nubilalis with any available methods. Biological and chemical methods are applied in the intervention control, but they do not protect maize plants fully against the pest (Bereś, 2006; Bereś & Lisowicz, 2005; Lisowicz ,2003).

One of the most effective methods to control the expansion of corn borers and to reduce the damage is to cultivate maize varieties resistant to its feeding (Baute et al., 2002).

The purpose of the research described here was to compare the susceptibility of selected cultivars of conventional maize to damage caused by O. nubilalis with transgenic varieties under the conditions of south-eastern Poland. Materials and Methods Initial research was carried out in 2005 in two locations in the area of Małopolskie (A) and Podkarpackie (B) provinces. Four varieties of maize were used: 2 cultivars of conventional maize (PR39H32 and PR39D81) and 2 transgenic ones (DKC3421YG and PR39D82 both expressing Cry1Ab, transformation event MON810). Plants were sown in random arrangements in four replications. The assessment of damage caused by Ostrinia nubilalis was carried out prior to crop harvest (BBCH 87) (Adamczewski and Matysiak 2002). For this purpose, 100 consecutive plants of each field plot were examined. The percentages of damaged plants were calculated, stalks broken below and above the cobs, as well as cobs bitten at the cob base were counted.

In 2006, the conventional varieties DKC3420 and PR39D81 and the same transgenic varieties as in 2005 were grown at locations A and B. The testing range was broadened by a third location (C) in the Lubelskie province, where the conventional cultivars DKC3420 and CLARICA and the transgenic counterparts DKC3421YG and BACILLA (expressing Cry1Ab, transformation event MON810) were grown. The damage to maize plants for all locations was evaluated at two time intervals: – 30th Aug.-5th Sept. 2006, when the plants were in the stage of waxy kernel maturity (BBCH 85), the percentages of plants and cobs damaged by caterpillars were calculated by inspecting 135 consecutive plants on each field plot. At location C, 100 consecutive plants in a row were inspected at four areas in the field. Additionally, by cutting 4 x 25 consecutive plants, the average number of holes and caterpillars in plants was calculated. – 1st -5th Oct. 2006, when maize plants were in the stage of full maturity (BBCH 87), the percentages of stalks broken below and above the cob, as well as cobs bitten at the cob base were calculated. In locations A and B, 135 consecutive plants on each field plot were evaluated, whereas in the location C, 100 consecutive plants in a row were inspected at four areas for each cultivar. The results concerning the number of holes and caterpillars were analyzed statistically and significant differences were determined by means of Student’s T test (Oktaba, 1976).

The effectiveness of transgenic cultivars regarding the reduction of plant damage caused by Ostrinia nubilalis was calculated in comparison to their conventional counterparts.

Results In 2005-2006 the meteorological conditions were good for the growth and feeding of Ostrinia nubilalis. Particularly favorable weather conditions were recorded in the dry and hot year 2006. In the first year of research, the highest harmfulness of corn borer was recorded in location B, where caterpillars damaged in average 52.4% of the plants of both conventional

33

cultivars. In both locations, in the transgenic cultivars only single plants had traces of feeding by the pest. The effectiveness of Bt cultivars in reducing the number of damaged plants in locations A an B was 97.8 to 100 %. Plants with the Bt gene were also found to be highly resistant to other damage which determines yield quantity and quality, such as: damaged cobs, cobs bitten at the base and stalks broken above and below cobs (Table 1).

Table 1. Susceptibility of selected maize cultivars to Ostrinia nubilalis feeding in 2005 (location A: Małopolskie and B: Podkarpackie).

% of broken stalks Variety % infested plants

below cob above cob

% of cobs gnawed at the

base

Yield [dt/ha]

Localization A B A B A B A B A B

PR39H32 25.5 55.5 7.7 12.8 17.0 39.9 4.0 12.8 119.4 115.2 DKC3421YG* 0.56 0.3 0.0 0.37 0.3 0.0 0.0 0.0 132.4 141.9 Efficacy [%] 97.8 99.4 100 97.1 98.2 100 100 100 – –

PR39D81 33.4 49.3 11.7 11.8 18.1 32.0 12.3 15.2 118.3 127.8

PR39D82* 0.7 0.0 0.0 0.0 0.7 0.0 0.0 0.0 128.2 129.3 Efficacy [%] 97.9 100 100 100 96.1 100 100 100 – –

* transgenic variety

In 2006 O. nubilalis was most harmful at location B. The conventional cultivar, PR39D81, was most susceptible to damage in locations A and B. Both transgenic cultivars showed high resistance to feeding by pest caterpillars (Table 2).

Table 2. Susceptibility of selected maize cultivars to Ostrinia nubilalis feeding in 2006 (location A: Małopolskie and and B: Podkarpackie).

% infested % of broken stalks Variety plants

cobs below cob above cob

% of cobs gnawed at the

base

Yield [dt/ha]

Localization A B A B A B A B A B A B

DKC 3420 31.7 43.4 4.8 23.2 5.3 6.3 6.2 11.1 1.0 2.0 122.7 113.5 DKC 3421YG* 0.3 0.1 0 0.1 0.0 0.1 0.0 0.0 0.0 0.0 123.0 132.6 Efficacy [%] 99.0 99.7 100 99.5 100 98.4 100 100 100 100 – –

PR39D81 44.6 68.6 4.0 33.0 11.9 19.5 8.2 14.0 2.5 2.0 110.0 86.1 PR39D82* 0.3 0.1 0.0 0.1 0.1 0.1 0.0 0.0 0.0 0.0 125.5 113.6 Efficacy [%] 99.3 99.8 100 99.6 99.1 99.4 100 100 100 100 – –

* transgenic variety

At location C, Ostrinia nubilalis damaged in average 34.1% of the plants for both conventional cultivars. DKC3420 was most susceptible to feeding by the pest. Both transgenic cultivars exhibited high resistance to damage ranging from 99.4 to 100% (Table 3). The number of damaged plants was low as only single traces of feeding (holes), but no live

34

caterpillars of Ostrinia nubilalis were found in the plants (Table 4). The statistical analysis at significance level 0.05 (Student’s t-test) found that the differences in numbers of holes and caterpillars between conventional and transgenic cultivars were statistically significant.

Table 3. Susceptibility of selected maize cultivars to Ostrinia nubilalis feeding in 2006 (location C: Lubelskie).

% of broken stalks Variety % infested plants

% infested cobs

below cob above cob

% of cobs gnawed at the base

Yield [dt/ha]

DKC3420 42.0 20.2 8.25 4.25 1.75 67.8 DKC3421YG* 0.25 0.0 0.0 0.0 0.0 85.1 Efficacy [%] 99.4 100 100 100 100 –

CLARICA 26.2 7.5 4.75 3.0 0.25 63.8 BACILLA* 0.0 0.0 0.0 0.0 0.0 81.2 Efficacy [%] 100 100 100 100 100 –

* transgenic variety Table 4. Number of holes and caterpillars in maize plants in 2006 (location C: Lubelskie).

Number of holes Number of caterpillars Total in

plant

Effec-tiveness

[%]

Variety

in stalk in cob

Total in plant

Effec-tiveness

[%]

in stalk in cob

DKC3420 0.82 0.21 1.03 – 0.72 0.22 0.94 – DKC3421YG* 0.01 0.0 0.01 99.0 0.0 0.0 0.0 100

LSD (0.05) 0.49 0.18 – – 0.39 0.19 – –

CLARICA 0.78 0.09 0.87 – 0.67 0.08 0.75 –

BACILLA* 0.0 0.0 0.0 100 0.0 0.0 0.0 100

LSD (0.05) 0.38 0.02 – – 0.38 0,04 – –

* transgenic variety

In 2005 - 2006 a significant rise in transgenic varieties grain yield, compared to conventional varieties, was achieved in all locations (A, B, C). Moreover, yield of Bt cultivar were characterized by higher healthiness. Conclusions 1. Transgenic cultivars exhibited high resistance to feeding by caterpillars of Ostrinia

nubilalis in comparison to their conventional counterparts. 2. Caterpillar damage was reduced by 97.8 – 100% in 2005 and by 99.0 – 100% in 2006 in

all locations with the use of transgenic cultivars. The percentage of other damage having impact on the crop height and quality was reduced as well.

3. The percentage of damaged plants was lower because fewer caterpillars fed in stalks and cobs of transgenic cultivars, and fewer holes were bitten by them.

35

4. The studied transgenic cultivars proved to be useful for cultivation in the soil and climatic conditions of Poland.

5. In the situation of high pressure by Ostrinia nubilalis for maize and the problems encountered in biological and chemical control, the use of transgenic cultivars may be an alternative to consider.

References Adamczewski, K. & Matysiak, K. 2002: Klucz do Określania Faz Rozwojowych Roślin

Jedno– i Dwuliściennych w skali BBCH [tłumaczenie i adaptacja – K. Adamczewski, K. Matysiak]. Instytut Ochrony Roślin, Państwowa Inspekcja Ochrony Roślin i Nasiennictwa: 20-21.

Bereś, P. 2006: Efekty chemicznego zwalczania omacnicy prosowianki (Ostrinia nubilalis Hbn.) w południowo-wschodniej Polsce w latach 2003-2005. Prog. Plant Protection/Post. Ochr. Roślin 46: 464-467.

Bereś, P. & Kaniuczak, Z. 2006: Omacnica prosowianka groźna dla kukurydzy. Agrotechnika, nr 5: 12-14.

Bereś; P. & Lisowicz, F. 2005: Przydatność kruszynka (Trichogramma spp.) w ochronie kukurydzy przed omacnicą prosowianką (Ostrinia nubilalis Hbn.) w gospodarstwach ekologicznych. Prog. Plant Protection/Post. Ochr. Roślin 45 (1): 47-51.

Baute, T.S., Sears, K.M. & Schaafsma, A.W. 2002: Use of transgenic Bacillus thuringiensis Berliner Corn Hybrids to determine the direct economic impact of the European corn borer (Lepidoptera: Crambidae) on field corn in Eastern Canada. J. Econ. Entomol. 95: 57-64.

Kania, C. 1961: Z badań nad omacnicą prosowianką – Pyrausta nubilalis (Hbn.) na kukurydzy w okolicach Wrocławia w latach 1956-1959. Pol. Pismo Entomol., Seria B 3-4: 165-181.

Kania, C. 1962: Szkodliwa entomofauna kukurydzy obserwowana w okolicach Wrocławia w latach 1956-1959 (cz.I). Pol. Pismo Entomol., Seria B, 1-2: 53-59.

Lisowicz, F. 1995: Omacnica prosowianka – groźny szkodnik kukurydzy. Ochrona Roślin, nr 5: 6-7.

Lisowicz, F. 2001: The occurrence of economically maize pests in south-eastern Poland. J. Plant Protection Res. 41: 250-255.

Lisowicz, F. 2003: The occurrence and the effects of European corn borer (Ostrinia nubilalis Hbn.) control on corn crop in Przeworsk region in 2001–2002. J. Plant Protection Res. 43: 399-403.

Lisowicz, F. & Tekiela, A. 2004: Szkodniki i choroby kukurydzy oraz ich zwalczanie. Technologia Produkcji Kukurydzy. Publishing by „Wieś Jutra”, Warsaw: 52-64.

Oktaba, W. 1976: Elementy statystyki matematycznej i metodyka doświadczalnictwa. PWN, Warsaw, 310 pp.

36

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 37-42

Baseline susceptibility of Helicoverpa armigera (Hübner) to Bt toxins Cry1Ac and Cry2Ab2 in West Africa Thierry Brevault1,2, Patrick Prudent1,3, Maurice Vaissayre1

1 CIRAD, UPR 102, Annual Cropping Systems, Montpellier, France (E-mail: maurice.vaissayre@cirad.fr); 2 IRAD, Cotton Programme, Garoua, Cameroon; 3 INRAB, CRA-Coton & Fibres, Cotonou, Benin Abstract: The susceptibility of African populations of the bollworm Helicoverpa armigera (Hübner) to Bt toxins is still unknown, and nothing has been published regarding resistance management measures required as a follow-up to the introduction of second-generation Bt cottons in Africa. CIRAD (Centre de Coopération Internationale en Recherche Agronomique pour le Développement) is currently building a model to help local authorities to develop a strategy for a sustainable use of Bt cotton in small farming systems. A key input factor for the model is the efficacy of Bt toxins against naïve populations of the bollworm. The purpose of the present study was to determine the susceptibility of bollworm strains collected in various cotton growing areas of Western and Central Africa to Cry1Ac and Cry2Ab as well as to their association in a 1/1 ratio. As expected according to their mode of action, Bt toxins had detrimental effects on H. armigera larvae, both in terms of mortality and larval growth. Despite some methodological difficulties in evaluating larval mortality following Bt toxin ingestion by H. armigera larvae, the LC50 as well as the GI50 values obtained in West Africa were quite similar to data published elsewhere in the Old World. Keywords: Bt Cotton, Cry toxins, Bioassays, Baselines, Helicoverpa armigera, Africa Introduction Bt cotton was introduced in the Old World (Australia, China, India) to manage the bollworm Helicoverpa armigera Hübner, whose populations became resistant to pyrethroid insecticides. In West Africa, Burkina Faso will be the first country to grow Bt cotton at a commercial scale, and there is no doubt that neighbouring cotton producing countries will adopt the technology straight after1. A refuge strategy to manage resistance is difficult to implement in the fragmented landscape of small scale farming systems in West Africa. To preserve efficacy of Bt cotton against bollworms, Burkinabe scientists, cooperating with Monsanto (Saint Louis, USA), have chosen to introduce one genetically engineered African cultivar expressing two toxins: Cry1Ac and Cry2Ab2.

Up to now, the susceptibility of bollworms2 to Bt toxins remains unknown in West Africa. CIRAD (Centre de Coopération Internationale en Recherche Agronomique pour le Développement) is currently developing a model to help local authorities in a proper implementation of a sustainable management of Bt cotton, (Vaissayre et al., 2006; Nibouche

1 Among the main cotton producting countries in West Africa, Burkina Faso is the first one to have officially announced the introduction of Bt cotton (planned initially for 2007, and postponed recently to 2009) after adopting a legal framework for the introduction of GM crops. 2 Not only H. armigera, but also Diparospis watersi Roths., Earias spp., Cryptophlebia leucotreta Meyr. and Pectinophora gossypiella Saund. are considered as harmful to cotton in West and Central Africa, with a variable pest status, according to the climatic zone.

37

38

et al., 2007). Key parameters are the efficacy of Bt toxins against H. armigera, considered as the key pest of cotton in West Africa, and the frequency of resistant individuals in naïve populations (we use the term "naïve" to make clear that this work is done on insect populations never exposed to Bt toxins before). The present study aimed to examine the susceptibility of various H. armigera strains to Cry1Ac and Cry2Ab2 toxins, either in terms of mortality, or, according to the specificity of Bt toxins, in terms of inhibition of larval growth. Bollworm strains were collected in various cotton growing areas of Western (Benin) and Central (Cameroon) Africa. Material and Methods Insect strains Bioassays were performed in 2006, from mid September to the end of December at IRAD (Institut de Recherche Agricole pour le Development) facilities in Garoua, Cameroon and from fall 2006 to early 2007 in INRAB (Institut National des Recherches Agricoles du Bénin) facilities in Bohicon, Benin.

Bollworm larvae were collected in cotton fields from 7 localities in Cameroon, 2 localities in Chad and a cotton growing village in North Eastern Nigeria, as representative samples of bollworm populations from dry savannas (Figure 1). Other strains were collected in various cotton growing regions of Benin, either on cotton (fall 2006) or on tomatoes (end 2006 for tomato 1, early 2007 for tomato 2) as representatives of populations from the coastal area. In each place, 100 to 200 late instars of H. armigera larvae were collected one by one along a 10 km transect, to avoid collecting larvae from a single progeny.

Larvae were reared on a semi-artificial diet in the laboratory. Adults obtained were placed in jars for egg laying. The bioassays were performed on the first instars (24 h after hatching) from F1 and F2 progeny. Toxins Two toxins were tested: Cry 1Ac (200 mg/g) was obtained as MVPII bioinsecticide from Dow Chemicals (Indianapolis, USA) and Cry 2Ab2 (6 mg/g) as lyophilized corn leaf powder from Monsanto Company (Chesterfield, USA). A toxin mixture (Cry1Ac + Cry2Ab2), 1/1 ratio, was also tested.

Dilutions of powder in distilled water led to seven doses : 0.01/ 0.04/ 0.13/ 0.45/ 1.60/ 5.63 and 20 µg/ml of diet for Cry1Ac toxin as well as 0.1/ 0.2/ 0.6/ 1.4/ 3.4/ 8.3 and 20 µg /ml of diet for Cry2Ab toxin, and a control. The semi synthetic diet used for bioassays was described by Giret and Couilloud (1986). When the temperature of the semi-synthetic diet was reduced to 45-50°C, the toxin suspensions were incorporated into the runny diet to obtain a volume of 150 ml. One ml of diet with the toxin was dispensed into each of the 24 cells of a plate, and 6 plates prepared for each dose of toxin. The diet was allowed to get cold under UV for 1 hour.

Bioassays were conducted by exposing neonates (< 24 h after hatching) to treated artificial diet. One bollworm larva was transferred into each cell, the cell closed with Parafilm

® and the plates maintained at a temperature of 25 ± 2°C, RH 60 ± 20% and a photoperiod of 14:10h. Each test (24 larvae) was repeated 3 to 6 fold.

39

Figure 1. Insect collection sites in Central Africa (Cameroon, Chad and Nigeria). Observations Mortality was recorded, if any, after 1, 2, 5 and 7 days. For the determination of the LC50, we considered not only dead larvae, but larvae not reaching the third instar after 7 days were also considered as dead ones.

According to Siegfried et al. (2000), the feeding disruption effect of Bt toxins was evaluated: the weight of the living larvae was recorded at 7 days, and compared with the control, to obtain a growth inhibition factor (GI50 = dose corresponding to 50% of larvae on treated diet reaching less than half the weight of larvae growing on the control diet).

Statistical analysis was performed using WinDL V2.0 software (CIRAD, 1999). Results and Discussion Data were obtained in Benin on strains collected from cotton in October, and from tomato later on (from December 2006 to March 2007). An increasing toxin concentration (Figure 2) led more quickly to 50% mortality in the population whose parents were collected on cotton than for the populations whose parents were collected on tomatoes. Such a trend disappeared for higher mortality levels, and the results obtained from various sampling dates on tomatoes were not significantly different. Toxic effects on insect larvae found at lower mortality levels could have been affected by the plant that hosted and fed the parent population.

100

Djalingo

Mokong

Sites

Tchatibali

Homé

Guider

Gaschiga

Tcholliré

Doba

Sorga

Gombe

CHAD

CAMEROON

NIGERIA Cotton area of Cameroon

0 50

40

Cry1Ac

0

20

40

60

80

100

0,00 0,00 0,01 0,02 0,06 0,19 0,45 0,85 1,32

log-dosis toxin

mor

talit

y (c

orre

cted

> L

2)

cottontomatotomato1tomato2

Figure 2. Larval mortality according to adult host plants and sampling date for Cy1Ac toxin.

In terms of growth inhibition (Figure 3), data are not significantly different, but a similar trend was observed, independently from the toxin used.

Cry1Ac

0,00

0,02

0,04

0,06

0,08

0,10

0,00 0,00 0,01 0,02 0,06 0,19 0,45 0,85 1,32

log-dosis toxin

mea

n w

eigh

t, 7d

-old

la

rvae

cottontomatotomato1tomato2

Figure 3. Weight decrease associated with adult host plant and sampling date for Cry1Ac toxin.

For insect strains collected in and around Cameroon, data obtained for LC50 (Figure 4) as

well as for GI50 (fig 5) are illustrating the natural variability in naïve populations. Cry 1Ac confirmed a high level of toxicity, expressed as direct mortality through LC50,

and a feeding disruption effect, expressed as a decrease or an arrest of feeding, followed by a loss of weight (illustrated by GI50 data). The mean value for LC50 was 1.1 µg/ml and ranged

41

from 0.2 to 2.28. Growth inhibition occurred for concentrations ten-fold lower: 0.15 µg/ml (0.05 to 0.38). Our results are in accordance with previous ones (Kranthi et al., 2001; Jalali et al., 2004) for Cry1Ac on H. armigera.

Cry 2Ab had a lower impact on both mortality and feeding. The mean value for LC50 was 2.14 µg/ml and ranged from 0.65 to 5.9, while growth inhibition occured for concentrations ten-fold lower: 0.26 µg/ml (0.04 to 0.55).

Associating both toxins in a 1/1 ratio lead to intermediate values for the LC50: 1.63 µg/ml, but remained similar to Cry2Ab alone for the GI50.

LC50 variability for Bt toxins

0

1

2

3

4

5

6

7

Djalingo

Doba

Gasch

iga

Guider

Homé

Mokon

gSorg

a

Tchati

bali

Tcholl

iré

Gombe

sampling points

LC50

Cry1AcCry2AbCry1+2

Figure 4. variability of LC50 data according to sampling place and toxins.

GI50 variability for Bt toxins

0

0,1

0,2

0,3

0,4

0,5

0,6

Djalingo

Doba

Gasch

iga

Guider

Homé

Mokon

gSorg

a

Tchati

bali

Tcholl

iré

Gombe

sampling points

GI5

0

Cry1AcCry2AbCry1+2

Figure 5. variability of GI50 data according to sampling place and toxins.

42

Acknowledgements We thank Monsanto and Dow Chemical Companies for providing Bt toxins, and the participation of INRAB and IRAD technicians in field collection and rearing of insects. This work is a part of a research project supported by a special grant from CIRAD. References Giret, M. & Couilloud, R. 1986: Remplacement de l'Agar-agar par un gélifiant à base de

carraghénate pour la confection de milieux nutritifs destinés à l'élevage des Lépidoptères. Coton & Fibres Tropicales 41: 131-133.

Jalali, S.K., Mohan, K.S., Singh, S.P., Manjuntah, T.M. & Lalitha, Y. 2004 : Baseline susceptibility of the old-world bollworm, Helicoverpa armigera (Lepidoptera: Noctuidae) populations from India to Bacillus thuringiensis Cry1Ac insecticidal protein. Crop Protection 23: 53-59.

Kranthi, K.R., Kranthi, S. & Wanjari, R.R. 2001: Baseline toxicity of Cry1A toxins to Helicoverpa armigera Hübner (Lepidoptera: Noctuidae) in India. International Journal of Pest Management 47: 141-145.

Nibouche, S., Guérard, N., Martin, P. & Vaissayre, M. 2007: Modelling the role of refuges for sustainable management of dual-gene Bt Cotton in West African smallholder farming systems. Crop Protection 26: 828-836.

Siegfried, B.D., Spencer, T. & Nearman, J. 2000: Baseline Susceptibility of the Corn Earworm (Lepidoptera: Noctuidae) to the Cry1Ab Toxin from Bacillus thuringiensis. Journal of Economic Entomology 93: 1265-1268.

Vaissayre, M., Martin, P. & Nibouche, S. 2006: Key factors for Bt cotton sustainability in smallholder farming: a modelling approach. IOBC/WPRC Bulletin 29(5): 183-186.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 43-49

Direct effects of Galanthus nivalis agglutinin (GNA) and avidin on the ladybird beetle Coccinella septempunctata Mukesh K. Dhillon1,2, Nora C. Lawo2, H.C. Sharma1, Jörg Romeis2 1International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India; 2Agroscope Reckenholz-Tänikon Research Station ART, Reckenholzstrasse 191, 8046, Zurich, Switzerland (E-mail: joerg.romeis@art.admin.ch) Abstract: Genes encoding Galanthus nivalis agglutinin (GNA) and avidin have been incorporated in several crops to enhance their resistance to a range of insect pests. The ladybird beetle, Coccinella septempunctata (L.) is an important predator of aphids and other soft-bodied insects in different crops. Thus, C. septempunctata is likely to ingest insecticidal proteins expressed by genetically modified (GM) plants either directly by feeding on pollen or through its prey. We have conducted two experiments to test the direct effects of GNA and avidin on a range of life-table parameters of C. septempunctata. The insecticidal proteins were provided dissolved in a 2M sucrose solution at a concentration of 1% (weight per volume). In the first experiment, neonate C. septempunctata larvae were fed either a pure sucrose solution (control) or a sucrose solution containing GNA or avidin. Every alternate day, predator larvae were fed exclusively with aphid prey. Ingestion of avidin resulted in a reduced C. septempunctata larval survival and adult emergence compared to the control group. It appeared that the predator larvae were more sensitive to GNA, expressed in a 100% mortality. In the second experiment first instar predator larvae were fed exclusively and continuously with pure sucrose solution (control) or sucrose solution containing GNA or avidin. While GNA reduced larval longevity significantly, avidin consumption caused no effect compared to the control. The results indicate that both GNA and avidin pose a hazard for C. septempunctata larvae under high-dose exposure conditions. To assess the risk that GNA- or avidin-expressing GM plants would pose to C. septempunctata, additional tests under more realistic exposure conditions would need to be conducted. Key words: Coccinella septempunctata, GNA, avidin, direct effects, non-target risk assessment Introduction Over the past two decades there has been a considerable progress in handling and introducing novel genes into crop plants to increase yields, improve nutrition, and impart resistance to biotic and abiotic stresses (Jouanin et al., 1998; Sharma et al., 2004). Insect-resistant transgenic crops expressing cry genes from the soil bacterium Bacillus thuringiensis (Bt) have been grown commercially since 1996 (Shelton et al., 2002) and appear to have little unintended effects on biological control organisms (Romeis et al., 2006). However, many other insecticidal proteins, e.g. the snowdrop lectin (Galanthus nivalis agglutinin; GNA) and avidin, have been introduced into a number of crop plants for resistance against coleopteran, lepidopteran, and homopteran insect pests (Bruins et al., 1991; Powell et al., 1995, Kramer et al., 2000; Markwick et al., 2001; Malone et al., 2002). Although GNA and avidin appear to be non toxic to mammals (Kramer et al., 2000), there are concerns about plants expressing these insecticidal proteins to cause unintended effects on non-target insects including natural enemies. Since both GNA and avidin have been found to affect pests of the order Coleoptera, it is likely that they also cause direct toxic effects on ladybird species that are important

43

44

predators of aphids and other soft-bodied insects in different agro-ecosystems. The present studies were conducted to assess these direct effects on Coccinella septempunctata (L.). Materials and Methods Insects Eggs of C. septempunctata were obtained from Katz Biotech Services, Germany, and kept at 22 ± 1ºC until hatching. A culture of Aphis craccivora Koch (Hemiptera: Aphididae) was maintained on broad bean (Vicia faba) at 24 ± 2ºC, 80 ± 5% r.h. with a 16h photoperiod. Insecticidal proteins Lyophilised GNA, isolated from snowdrop bulbs, was obtained from E.J.M. Van Damme (Ghent University, Belgium; Van Damme et al., 1995). Recombinant avidin, purified from transgenic corn (≥ 12 units/mg protein) was purchased from Sigma-Aldrich, Switzerland. Proteins were dissolved in a 2M sucrose solution at a concentration of 1% (weight per volume, w:v). The solutions were stored at -20ºC until use. Experimental conditions All bioassays were conducted in a climate chamber at 24 ± 1°C, 80 ± 5% r.h. with a 16h photoperiod. Bioassays were conducted with freshly hatched C. septempunctata larvae, which were kept individually in a Petri dish (5cm diameter, 1cm high) the lid of which contained a hole covered with fine-mesh netting for ventilation. Direct effects of GNA and avidin on different life-table parameters of C. septempunctata Individual, freshly hatched (neonate) C. septempunctata larvae received one of the following three food solutions: (i) pure 2M sucrose solution (“sucrose + aphids”), (ii) 2M sucrose solution containing 1% GNA (“GNA + aphids”), or (iii) 2M sucrose solution containing 1% avidin (“avidin + aphids”). After every 24h of feeding on the food solution, the C. septempunctata larvae were provided ad libitum with A. craccivora (mixed stages). Subsequently, the predator larvae were fed with the food solution or aphids every alternate day. The five different C. septempunctata instars received 0.5, 1, 3, 4, or 5μl of the food solution, respectively, to ensure ad libitum access to food. As a control treatment, predator larvae were fed exclusively on A. craccivora throughout their larval development (“aphids”). A total of 60 C. septempunctata larvae were tested per treatment in a complete randomised design. The following life-table parameters were recorded: larval and pupal developing time, larval survival, adult emergence, male and female weights.

The GNA treatment (“GNA + aphids”) was excluded from the statistical analyses since all larvae died during the observation period. For all life-table parameters two pairwise comparisons were made: (i) “sucrose + aphids” vs. “aphids” and “sucrose + aphids” vs. “avidin + aphids”. Bonferroni correction was applied leading to an adjusted α = 0.025. Data on larval and pupal developing time was analysed using Mann-Whitney-U test, larval survival and adult emergence by Fisher’s exact test, male and female weights by Student’s t-test.

Effects of GNA and avidin on the longevity of neonate C. septempunctata Neonate C. septempunctata were exclusively provided with one of the following food solutions: (i) pure 2M sucrose solution, (ii) 2M sucrose solution containing 1% GNA (w:v), (iii) 2M sucrose solution containing 1% avidin (w:v), or (iv) water. The last treatment served as a control to ensure that the predator larvae had actually ingested the provided food solutions and was excluded from the subsequent statistical analysis. A total of 30 C.

45

septempunctata larvae were used per treatment in a completely randomised design. Longevity of the C. septempunctata larvae was recorded twice a day.

Longevity of C. septempunctata larvae was compared separately between the pure sucrose solution and the sucrose solution containing on of the insecticidal proteins by Mann-Whitney-U test with Bonferroni correction (adjusted α = 0.025). Results Direct effects of GNA and avidin on different life-table parameters of C. septempunctata A significant reduction in larval development time, a prolonged pupal development time and an increased adult weight were observed when C. septempunctata larvae were exclusively fed with “aphids” compared to those fed on “sucrose + aphids” (Table 1). However, no significant differences were detected for larval survival and adult emergence. Larvae that consumed “avidin + aphids” suffered a significantly increased mortality and had a reduced adult emergence when compared to “sucrose + aphid” fed insects (Table 1). All other parameters were not affected by avidin. All C. septempunctata larvae feeding on “GNA + aphids” died. Table 1. Direct effects of GNA and avidin on different life-table parameters of Coccinella septempunctata (mean ± SE).

a The “GNA + aphids” treatment excluded from the statistical analysis.

Treatments Larval dev. time

(days)

Larval survival

(%)

Pupal dev. time

(days)

Adult

emergence (%)

Male weight

(mg)

Female weight

(mg)

(i) Aphids 11.0±0.17 87.9±4.31 4.7±0.06 74.1±5.80 22.65±0.67 26.43±0.52

(ii) Sucrose

+ aphids 13.9±0.21 83.1±4.93 4.4±0.07 74.6±5.72 19.86±0.52 24.82±0.40

(iii) Avidin

+ aphids 14.4±0.42 30.0±5.97 4.4±0.13 23.3±5.51 18.01±0.97 23.83±0.65

(iv) GNA

+ aphids a - b 0.0 - 0.0 - -

(i) vs. (ii) c P = <0.0001 P = 0.601 P = 0.0027 P = 1.000 P = 0.003 P = 0.019

(ii) vs. (iii) c P = 0.210 P < 0.0001 P = 0.850 P < 0.0001 P = 0.110 P = 0.220

b No larvae survived. c The P-values are based on α-levels adjusted for two pairwise comparisons (α = 0.025). Effects of GNA and avidin on the longevity of neonate C. septempunctata There were significant differences in the longevity of C. septempunctata larvae on different food solutions (Figure 1). Neonates survived significantly longer on a pure sucrose solution compared to those feeding on a sucrose solution containing GNA. In contrast, consumption of avidin had no effect on larval longevity. C. septempunctata larvae survived only for about 1.5

46

days on water. The fact that larvae lived much longer in the other treatments indicates that they had consumed the provided sucrose solutions.

0

1

2

3

4

5

6

7

Water Sucrose Sucrose+ avidin

Sucrose+ GNA

Lar

vall

onge

vity

(day

s)

P < 0.001

P = 0.72

0

1

2

3

4

5

6

7

Water Sucrose Sucrose+ avidin

Sucrose+ GNA

Lar

vall

onge

vity

(day

s)

P < 0.001

P = 0.72

Figure 1. Mean longevity (± SE) of Coccinella septempunctata larvae on different food solutions (n = 30). Discussion None of the C. septempunctata larvae fed with “GNA + aphids” survived until pupation. Also larvae fed exclusively with sucrose solution containing GNA had a significantly shorter longevity than sucrose fed larvae. These results confirm previous studies that have shown direct toxic effects of GNA on a number of insect predators and parasitoids. For example, earlier studies reported a reduction in the longevity of larvae of the aphid predators Chrysoperla carnea (Stephens) (Neuroptera: Chrysopidae), Adalia bipunctata (L.) (Coleoptera: Coccinelidae) and C. septempunctata when ingesting a 1% GNA-sucrose solution (Hogervorst et al., 2006; Lawo & Romeis, 2008). Birch et al. (1999) reported that the fecundity, egg viability and longevity of ladybirds significantly decreased when feeding on aphids reared on GNA-transgenic potatoes. GNA appears to bind to mid-gut epithelial cells of ladybird larvae (Hogervorst et al., 2006), which might cause irreversible damage, and could explain the larval mortality observed in our experiments. Further on, studies with aphid parasitoids such as Aphelinus abdominalis (Hymenoptera: Aphelinidae) and Aphidius ervi (Hymenoptera: Aphidiidae) report sublethal effects by GNA when provided in form of GNA-fed aphid hosts (Couty et al., 2001a, b). A reduced longevity and fecundity was reported for the aphid parasitoids, Aphidius colemani (Hymenoptera: Braconidae) and the caterpillar parasitoid Eulophus pennicornis (Hymenoptera: Eulophidae) when feeding a GNA-sucrose solution (Romeis et al., 2003; Bell et al., 2004).

Consumption of “avidin + aphids” caused an increased mortality, a prolongation in larval developing time, and a decreased adult emergence in C. septempunctata. It is known that avidin as a growth inhibitor binds strongly to the vitamin biotin, resulting in biotin deficiency, which in turn leads to stunted insect growth and mortality. An addition of biotin to the avidin treated diet was shown to prevent the detrimental effects of the vitamin binding protein and allows normal growth of Lepidoptera (Morgan et al., 1993). In our second experiment,

47

longevity of C. septempunctata larvae was not affected when the predators were fed exclusively with avidin dissolved in a sucrose solution. This lack of an avidin effect could be due to the fact that the predators did not receive any biotin containing prey. So far, direct negative effects of avidin have been reported for arthropods belonging to the orders of Coleoptera, Lepidoptera, Acaridae, Orthoptera and Neuroptera at concentrations similar to the one used in our study (Levinson et al., 1992; Morgan et al., 1993; Kramer et al., 2000; Marwick et al., 2001; Malone et al., 2002; Zhu et al., 2005; Lawo & Romeis, 2008).

While our experiments reveal that GNA and avidin, at the high concentration provided, pose a hazard to C. septempunctata larvae, this does not imply that transgenic plants expressing these proteins would pose a risk to this important predator in the field. To assess the risk one would need to know plant expression levels of the insecticidal proteins and at what level the predator would be exposed.

The expression levels of GNA in transgenic crops that might be released in the field are difficult to forecast. Currently technologies are being developed to concentrate the expression of this lectin to the phloem sap to target sap sucking pests by using phloem specific promoters (Rao et al., 1998; Foissac et al., 2000). A minimum of 0.1% GNA in the diet is required to achieve a detectable impact on aphids (Down et al., 1996; Couty et al., 2001a), a concentration that has no or marginal effects on A. bipunctata larvae (Birch et al., 1999; Down et al., 2000). However, to fight aphids sufficiently, GNA-expressing plants need to contain higher GNA levels in the phloem sap, which appears to harm beneficial insects such as C. septempunctata and C. carnea (Lawo & Romeis, 2008). Studies with avidin-expressing GM maize plants have revealed that expression levels of about 300 ppm would be required to provide good control of storage insect pests while individual transformed plants contained up to 2500 ppm of avidin (Kramer et al., 2000). Compared to this, our 1% concentration (10,000 ppm) appears to be rather high.

In case that GNA- or avidin-expressing GM plants become available, additional studies would need to be conducted to assess the risk to C. septempunctata under more realistic exposure conditions. Acknowledgements This study has been conducted within the Indo-Swiss Collaboration in Biotechnology (ISCB) http://iscb.epfl.ch. Funding by Swiss Agency for Development and Cooperation (SDC), Berne, Switzerland, and the Department of Biotechnology (DBT), New Delhi, India, is gratefully acknowledged. References

Bell, H.A., Kirkbride-Smith, A.E., Marris, G.C., Edwards, J.P. & Gatehouse, A.M.R. 2004:

Oral toxicity and impact on fecundity of three insecticidal proteins on the gregarious ectoparasitoid Eulophus pennicornis (Hymenoptera: Eulophidae). Agricultural and Forest Entomology 6: 215-222.

Birch, A.N.E., Geoghegan, I.E., Majerus, M.E.N., McNicol, J.W., Hackett, C.A., Gatehouse, A.M.R. & Gatehouse, J.A. 1999: Tri-trophic interactions involving pest aphids, predatory 2-spot ladybirds and transgenic potatoes expressing snowdrop lectin for aphid resistance. Molecular Breeding 5: 75-83.

Bruins, B.G., Scharloo, W. & Thorig, G.E.W. 1991: The harmful effect of light on Drosophila is diet-dependent. Insect Biochemistry 21: 535-539.

48

Couty, A., de la Vina, G., Clark, S.J., Kaiser, L., Pham-Delegue, M.H. & Poppy, G.M. 2001a: Direct and indirect sublethal effects of Galanthus nivalis agglutinin (GNA) on the development of a potato-aphid parasitoid, Aphelinus abdominalis (Hymenoptera: Aphelinidae). Journal of Insect Physiology 47: 553-561.

Couty, A., Down, R.E., Gatehouse, A.M.R., Kaiser, L., Pham-Delegue, M.H. & Poppy, G.M. 2001b: Effects of artificial diet containing GNA and GNA-expressing potatoes on the development of the aphid parasitoid Aphidius ervi Haliday (Hymenoptera: Aphidiidae). Journal of Insect Physiology 47: 1357-1366.

Down, R.E., Ford, L., Woodhouse, S.D., Raemaekers, R.J.M., Leitch, B., Gatehouse, J.A. & Gatehouse, A.M.R. 2000: Snowdrop lectin (GNA) has no acute toxic effects on a beneficial insect predator, the 2-spot ladybird (Adalia bipunctata L.). Journal of Insect Physiology 46: 379-391.

Down, R.E., Gatehouse, A.M.R., Hamilton, W.D.O. & Gatehouse, J.A. 1996: Snowdrop lectin inhibits development and decreases fecundity of the glasshouse potato aphid (Aulacorthum solani) when administered in vitro and via transgenic plants both in laboratory and glasshouse trials. Journal of Insect Physiology 42: 1035-1045.

Foissac, X., Loc, N.T., Christou, P., Gatehouse, A.M.R. & Gatehouse, J.A. 2000: Resistance to green leafhopper (Nephotettix virescens) and brown planthopper (Nilaparvata lugens) in transgenic rice expressing snowdrop lectin (Galanthus nivalis agglutinin; GNA). Journal of Insect Physiology 46: 573-583.

Hogervorst, P.A.M., Ferry, N., Gatehouse, A.M.R., Wäckers, F.L. & Romeis, J. 2006: Direct effects of snowdrop lectin (GNA) on larvae of three aphid predators and fate of GNA after ingestion. Journal of Insect Physiology 52: 614-624.

Jouanin, L., Bonadé-Bottino, M., Girard, C., Morrot, G. & Giband, M. 1998: Transgenic plants for insect resistance. Plant Science 131: 1-11.

Kramer, K.J., Morgan, T.D., Throne, J.E., Dowell, F.E., Bailey, M. & Howard, J.A. 2000: Transgenic avidin maize is resistant to storage insect pests. Nature Biotechnology 18: 670-674.

Lawo, N.C. & Romeis, J. 2008: Assessing the utilization of a carbohydrate food source and the impact of insecticidal proteins on larvae of the green lacewing, Chrysoperla carnea. Biological Control 44: 389-398.

Levinson, H.Z., Levinson, A.R. & Offenberger, M. 1992: Effect of dietary antagonists and corresponding nutrients on growth and reproduction of the flour mite (Acarus siro L.). Experientia 48: 721-729.

Malone, L.A., Burgess, E.P.J., Mercer, C.F., Christeller, J.T., Lester, M.T., Murray, C., Phung, M.M., Philip, B.A., Tregidga, E.L. & Todd, J.H. 2002: Effects of Biotin-binding proteins on eight species of pasture invertebrates. New Zealand Plant Protection 55: 411-415.

Markwick, N.P., Christeller, J.T., Dochterty, L.C. & Lilley, C.M. 2001: Insecticidal activity of avidin and streptavidin against four species of pest Lepidoptera. Entomologia Experimentalis et Applicata 98: 59-66.

Morgan, T.D., Oppert, B., Czapla, T.H. & Kramer, K.J. 1993: Avidin and streptavidin as insecticidal and growth inhibitory dietary proteins. Entomologia Experimentalis et Applicata 69: 97-108.

Powell, K.S., Gatehouse, A.M.R., Hilder, V.A. & Gatehouse, J.A. 1995: Antifeedant effects of plant-lectins and an enzyme on the adult stage of the rice brown planthopper, Nilaparvata lugens. Entomologia Experimentalis et Applicata 75: 51-59.

49

Rao, K.V., Rathore, K.S., Hodges, T.K., Fu, X., Stoger, E., Sudhakar, D., Williams, S., Christou, P., Bharathi, M., Bown, D.P., Powell, K.S., Spence, J., Gatehouse, A.M.R. & Gatehouse, J.A. 1998: Expression of snowdrop lectin (GNA) in transgenic rice plants confers resistance to rice brown planthopper. The Plant Journal 15: 469-477.

Romeis, J., Babendreier, D. & Wäckers, F.L. 2003: Consumption of snowdrop lectin (Galanthus nivalis agglutinin) causes direct effects on adult parasitic wasps. Oecologia 134: 528-536.

Romeis, J., Meissle, M. & Bigler, F. 2006: Transgenic crops expressing Bacillus thuringiensis toxins and biological control. Nature Biotechnology 24: 63-71.

Sharma, H.C., Sharma, K.K., Seetharama, N. & Crouch, J.H. 2004: Genetic engineering of crops for insect control: effectiveness and strategies for gene deployment. CRC Reviews in Plant Sciences 23: 47-72.

Shelton, A.M., Zhao, J.Z. & Roush, R.T. 2002: Economic, ecological, food safety, and social consequences of the deployment of Bt transgenic plants. Annual Review of Entomology 47: 845-881.

Van Damme, E.J.M., Smeets, K. & Peumans, W.J. 1995: The mannose-binding monocot lectins and their genes. Pages 59-80. In: Lectins: Biomedical Prospectives (Pusztai, A. and Bardocz, S., eds.). Taylor & Francis Ltd., London, UK.

Zhu, Y.C., Adamczyk, J.J. & West, S. 2005: Avidin, a potential biopesticide and synergist to Bacillus thuringiensis toxin against field crop insects. Journal of Economic Entomology 98: 1566-1571.

50

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 51-58

Validation of some techniques used in the evaluation of GM plant effects on tri–trophic interactions Julia Górecka, Zbigniew T. Dąbrowski, Monika Godzina, Karolina Kubis Department of Applied Entomology, Warsaw Agricultural University, Nowoursynowska St. 159, 02 – 776 Warsaw, Poland (E-mail: juliag2@wp.pl) Abstract: Our greenhouse and laboratory experiments initiated in 2005 have included DKc307, a maize cultivar expressing the cry1Ab gene (transformation event MON 810; Bt maize) as the GM reference crop and cv. Monumental, its near-isoline, as representatives of the first trophic level and: (a) the Mediterranean flour moth Ephestia kuehniella Zell. and its parasitoid - Venturia canescens; (b) Rhopalosiphum padi L. and its parasitoid Aphidius colemani (Viereck) as the second and third trophic level, respectively. Because our preliminary qualitative chemical analysis indicated some level of Cry1Ab toxin in ground DKc307 kernels, we have chosen E. kuehniella as phytophagous flour pest and its parasitoid -Venturia canescens. No significant differences in larval survival of E. kuehniella on Bt maize flour in comparison to the control cultivar was observed, however, the average weight of larvae reared on Bt maize flour was significantly lower. This effect did not affect the level of parasitism by V. canescens. In spite of our conclusion that V. canescens should not be used as a sensitive indicator of food quality change of its host, the results clearly showed that grain and flour made of Bt maize should be less infested by the moth larvae during their storage and the role of V. canescens as a parasitoid should not be disturbed. It was confirmed that the Bt maize did not cause a toxic effect on R. padi. In addition, aphids developed higher populations on the transgenic plants (significant only in the winter experiment), both in the winter and summer greenhouse experiments. Higher parasitization by A. colemani was observed on Bt maize in the experiments conducted during the winter. This result could not, however, be cofirmed during the summer experiments, indicating a seasonal effect on the tri-trophic relations. Key words: GM maize, MON 810, non-target organisms, tri-trophic relations, Ephestia kuehniella, Venturia canescens, Rhopalosiphum padi, Aphidius colemani Introduction Despite the fact that genetically modified (GM) crops are currently not allowed to be grown commercially in Poland, studies on the potential effects of GM plants on the environment should be carried out. It can not be excluded that the pressure from the farmers’ community may change the present restrictive position of the Government in Poland. Farmers have already witnessed significant advantages of growing insect-resistant Bt maize in southern Poland, where pest infestation (stem borers) of conventional susceptible cultivars can reach up to 70% in some fields. At the same time, various anti-GM groups are expressing their anxiety on potential negative unintended effects of GM crops on the environment and non-target organisms (NTOs) in particular.

Investigations of GM crop effects on NTOs are conducted since many years with financial support from EU programmes and national governments in the majority of the EU countries. There is presently consensus that the assessment of GM plant effects on NTOs should follow a step-wise (tired) approach. In the case of insect-resistant GM crops, early tier (laboratory) tests should be conducted to determine whether an organism is susceptible to the

51

52

expressed toxin under worse–case exposure conditions, conducted under standardized and repeatable conditions (Hill & Sendashonga, 2003; Dutton et al., 2003; Garcia-Alonso et al., 2006). If risks have been identified or can not be ruled out, additional (higher tier) tests should follow exposing NTOs to the toxin under more realistic conditions. Different techniques were used to measure toxin effects of GM plants on natural enemies under laboratory conditions: (a) via prey organisms previously fed on GM plants; (b) via a toxin incorporated into artificial diet; (c) via direct feeding on GM plants. It shall be noted that some scientists have questioned the validity of laboratory tests since they do not reflect realistic ecological scenarios by providing no prey choice and no additional stress factors (Lövei & Arpaia, 2005).

According to Romeis et al. (2006) the effects of Bt plants on hymenopterous parasitoids developing on herbivores reared on Bt-transgenic plants have been investigated in ten studies. The majority of cases were related to parasitoids of Lepidoptera species. Effects of maize and cotton cultivars expressing Cry toxins on mortality, development, weight or longevity of parasitoids were observed in cases where Bt-susceptible lepidopteran herbivores were used as hosts (e.g. Meissle et al., 2004; Bernal et al., 2002). The experiments have confirmed that host quality of the phytophagous species was most likely the cause of the negative effects on the parasitoids.

We have conducted greenhouse experiments to investigate the impact of Bt maize on tri-trophic interactions. The experimental plants included a maize cultivar expressing the cry1Ab gene (event MON 810, DKc307) and cv. Monumental, the corresponding near-isoline. The impact of the plants on the following insect systems (herbivore and natural enemy) were studied: (a) Mediterranean flour moth, Ephestia kuehniella Zeller (Lepidoptera: Pyralidae) and its parasitoid Venturia canescens Gravenhorst (Hymenoptera: Ichneumonidae); (b) the bird cherry-oat aphid, Rhopalosiphum padi L. (Hemiptera: Aphidae), a common phytophagous species of cereals in Poland, and the parasitoid Aphidius colemani (Viereck) (Hymenoptera: Braconidae). A. colemani is an exotic parasitoid commonly used to control aphids in greenhouses in Europe. From a technical point of view, the availability and amenability of a species is an important factor in increasing the reproducibility of the tests in different laboratories (Dutton et al., 2003; Dąbrowski & Górecka 2006).

Material and Methods Mediterranean flour moth and its parasitoid Venturia canescens Insects used in the experiment originated from colonies reared at the Department of Applied Entomology at Warsaw Agricultural University. First instar larvae of E. kuehniella were obtained from eggs collected in oviposition containers. Twenty neonates were transferred into glass jars filled with maize flour, 5 jars were treated as one replicate. The flour has been prepared by grinding the kernels of the transgenic maize cultivar and its non-transformed control. On each flour type a total of 1000 larvae were used (10 replications, 100 larvae per replication). After 3 weeks the larvae were weighted individually. The number of adults obtained from the larvae stock were counted after 4 weeks to estimate the mortality of the E. kuehniella larvae. The recorded parameters were: larvae body weight and number of adults emerged.

To examine the influence of host feeding on the parasitic wasp Venturia canescens, third instar E. kuehniella larvae were placed into little pockets made of polyurethane miller mesh with silicone hot glue on its selvage to prohibit escape of the larvae. Fifty larvae of E. kuehniella were transferred into one pocket and put into Plexiglas insectaries. V. canescens females were released into the insectaries at a rate of 4 females per 50 moth larvae. The miller

53

mesh allowed the wasps to find and parasitize the host larvae. Glass containers filled with a honey-water solution were kept in each insectary to provided food to the adult wasps. After one week, E. kuehniella larvae were transferred into glass jars in the environmental chamber. The number of adult wasps that emerged from the moth larvae was counted after 4 weeks. The bird cherry – oat aphid and parasitic wasp Aphidius colemani relations. Aphid colonies were maintained in the Experimental Greenhouse Center of the Faculty of Horticulture and Landscape Architecture, Ursynów. A stock colony of A. colemani was supplied by Rol – Eko sp. Company. Plants used in the experiment were grown in plastic pots located in two separate shelves at an average temperature of 25°C and 70% relative humidity, under a 16:8 h photoperiod regime. In spite of active ventilation system, the temperature in the greenhouse increased during the hot summer days of the year 2006. The experiment was conducted in the winter and in the summer of 2006.

Ten adult wingless aphid females were transferred onto each of ten maize seedlings (4–5 leaf growth stage) per variety in three replications. The number of progeny that developed on plants was counted three times at two week intervals.

After 8 weeks, plants were put into large cages covered by insect proof fine mesh (5 plants per cage). Into each cage 125 aphid mummies carrying pupae of A. colemani were transferred. To provide a food source for adult parasitoids, a glass Petri dish with a honey-water solution was provided inside the cage. After 3 weeks the number of emerged adults of A. colemani was estimated by counting the number of mummies with emerging holes. Statistical analyses In both experiments data was analyzed by one–way analysis of variance (Kruskal -Wallis test) by using STATGRAPHICS Plus 4.1 statistical package. Results Mediterranean flour moth and its parasitoid wasp Venturia canescens There was no significant difference in E. kuehniella larval survival on MON 810 flour in comparison to the flour of the corresponding non-transformed cultivar (Figure 1). However, the average weight of larvae after three weeks of feeding on Bt maize flour was significantly reduced (Figure 2). In addition, larvae reared on MON 810 flour required about 2 weeks more to develop and only few (12%) larvae pupated on Bt flour.

0123456789

10

Monumenta l M ON 810

Maize cultivar

Mea

n E

. kue

hnie

lla la

rvae

m

orta

lity

aa

Figure 1. Mean (± SE) Ephestia kuhniella larvae mortality on non-Bt and Bt (MON 810) maize flour (n =10) (p= 0.548; F=0.37).

54

0 5 10 15

Monumental

MON 810M

aize

cul

tivar

Mean E. kuehniella larval body weight

Figure 2. Influence of food type on Ephestia kuehniella on mean (± SE) larvae body weight (n=10) (p<0.0001; F=1543.94).

The differences in larval weight had no impact on the level of parasitisation by V. canescens. Four weeks after putting the wasps into the cage with E. kuehniella larvae, the percentage of parasitized caterpillars was 14,5% for moths feeding on flour made of Monumental kernels and 14,7% for larvae feeding on MON 810 cultivar (Figure 3).

Figure 3. Mean (± SE) parasitation of E. kuehniella by Venturia canescens on different maize flour (n=10) (p=0.950; F=0.00). The bird cherry – oat aphid and its parasitoid Aphidius colemani During both the winter and the summer experiment distinct differences in R. padi population development were observed. In the winter experiment the number of aphid progeny did not differ between Bt and control maize at the first count after two weeks but was higher on MON

02468

1012141618

Monumental MON 810

Maize cultivar

Perc

enta

ge o

f par

asite

d E.

ku

ehni

ella

larv

ae

20

a

b

aa

55

810 maize plants at the later sample dates (Figure 4). After 6 weeks, the R. padi population reached an average level of 258 aphids/plant on Bt maize in comparison to only 186 on its isoline (Figure 4). Similar differences in the R. padi population were observed in the experiment conducted during the summer season, but the population broke down between the second and fourth week because of the rapid temperature increase. After this period, the aphid population reached an average level of 600 aphids/plant on Bt maize and 459 aphids/plant on the isoline (Figure 5).

0

50

100

150

200

250

300

350

2 4 6

Week

Aver

age

num

ber o

f aph

ids

on

plan

ts.

MonumentalMON 810

b

a baa a Figure 4. Average (± SE) number of aphids on Bt and non-Bt maize plants – winter experiment (n=30, P: week 2 = 0.13, week 4 = 0.05, week 6 = 0.04)

0100200300400

500600700800900

2 4 6

Week

Ave

rage

num

ber o

f aph

ids

onpl

ants Monumental

MON 810

a a a

a

a

b

Figure 5. Average (± SE) number of aphids on Bt and non-Bt maize plants – summer experiment (n=30, P: week 2= 0.37; week 4=0.05; week 6= 0.04).

During the winter experiment a significantly higher percentage of parasitisation by A. colemani was noted for R. padi population developing on Bt maize plants (Figure 6). The mummification level of aphids indicated 3-fold increase (in comparison to introduced

56

population) of A. colemani population on Monumental and only 1.5-fold on the transgenic plants.

0 5 10 15 20

MON 810

Monumental

Mai

ze c

ultiv

ar

Percentage of parasited aphids

a

b

Figure 6. Mean (± SE) parasitation of Rhopalosiphum padi aphids feeding on Bt and non-Bt maize by Aphidius colemani – winter experiment (n = 30, p = 0.002)

However, this was not observed in the summer season where aphid parasitism levels on Bt and non-Bt maize were similar (Figure 7)

0 5 10 15 20 25 30

MON 810

Monumental

Mai

ze c

ultiv

ar

Percentage of parasited aphids

a

a

Figure 7. Mean (± SE) parasitation of Rhopalosiphum padi aphids feeding on Bt and non-Bt maize by Aphidius colemani – summer experiment (n =30, P = 0.37) Discussion Our laboratory experiments carried out under controlled conditions confirmed our preliminary observations that the bulk seeds of Bt maize (MON 810) were less infested by E. kuehniella than those from the corresponding non-transformed cultivar. Even so Bt maize flour did not have a significant effect on larval survival it caused a reduction in larval weight by 48.3%.

57

Furthermore the pupation rate was decreased on Bt maize flour. This Bt maize effect, however, did not affect the level of parasitasation by V. canescens. Thus, the parasitoid appears to be relatively insensitive to changes in the host’s quality. Our results clearly showed that grain and flour made of MON 810 during their storage should be less infested by the moth larvae without disturbing the parasitoid V. canescens.

The experimental results showing that Bt maize provides better conditions for R. padi population development when compared to the corresponding non-transformed maize cultivar may suggest that this phenomenon is specific to the original parent maize cultivar. Differences in the biochemical composition of Bt and non-Bt maize lines were found for lignin content, indicating some unexpected phenotypic changes in biochemical composition of Bt maize tissues (Saxena & Stotzky, 2001) while some authors carrying out field experiments did not observed a higher aphid population on Bt maize lines, others did. Lumbierrres et al. (2004) noted a higher density of R. padi, particularly alates and young nymphs in Bt plots at very young maize development stages, corresponding to the settlement period, in a 3 years field studies. The development and pre-reproductive times of the offspring of the first generation of alatae were shorter and the intrinsic rate of natural increase (rm) higher when aphids fed on Bt maize. However no differences on the aphid population parameters were found by the authors among the offspring of apterous aphids maintained on Bt or non-Bt maize for several generations (Lumbierres et al., 2004). Their results are in agreement with those reported by Lozzia et al. (1998) and Lozzia (1999) in the sense that any differential effect that appears between Bt and non-Bt maize disappears when the whole season was considered. Finally, the results obtained by these authors suggest, that no economic effects on maize crops should be expected.

Lathan & Wilson (2006), based on a large scale survey of the scientific literature, expressed the opinion that unexpected phenotypic consequences are common in transgenic plants. The authors, however, emphasized that most unexpected phenotypes reported had no known biosafety implications.

Bourguet et al. (2002) in their field studies carried out in two locations did not find significant differences between numbers of Metapolophium dirhodum (Walker), R. padi L. and Sitobion avenae (F.) on cv. Elgina (Monsanto hybrid MON 810) and cv. Cecilia (non-Bt maize). Under Southern Bohemia (Czech Republic) conditions, R. padi or M. dirhodum apparently preferred either the Bt or the non-Bt maize, but statistical evaluation of the data collected over the season revealed that these differences were insignificant (Sehnal et al., 2004). However, under Spanish field conditions, aphids of three species (R. padi; S. avenae and M. dirhodum) were in general more abundant on Bt maize (event 176) than on non-Bt control cultivar (Eizaguirre at al., 2006). Conclusions • Bt maize should be less infested by the storage pest E. kuehniella. • V. canescens did not respond to changes in host’s quality after feeding on Bt maize. • There was no clear result concerning the influence of Bt maize feeding of its host on A.

colemani. • The effect of season on the tri–trophic relations R. padi – A colemani indicates that the

one season evaluation is inadequate for proper estimation of GM plants on tri – trophic relationships.

58

Acknowledgements We wish to thank Monsanto for providing us with transgenic maize seeds. This study was partly supported by project P06R 05628 provided by the Ministry of Science and Higher Education of Poland.

References Bernal, J.S., Griset, J.G. & Gillogly, P.O. 2002: Impacts of developing on Bt maize-

intoxicated hosts on fitness parameters of a stem borer parasitoid. J. Entomol. Sci. 37: 27-40.

Bourguet, D., Chaufaux, J., Micoud, A., Delos, M., Naibo, B., Bombarde, F., Marque, G., Eychenne, N. & Pagliari, C. 2002: Ostrinia nubilalis parasitism and the field abundance of non-target insects in transgenic Bacillus thuringensis corn (Zea mays). Environ. Biosafety Res. 1: 49-60.

Dąbrowski, Z.T. & Górecka, J. 2006: [Methodology of risk assessment of transgenic crops expressing insecticidal toxins]. Prog. Pl. Protection/Post. Ochr. Roślin 46(1): 180-188.

Dutton, A., Romeis, J. & Bigler, F. 2003: Assessing the risks of insect resistant transgenic plants on entomophagous arthropods: Bt-maize expressing Cry 1Ab as a case study. BioControl 48: 611-636.

Eizaguirre, M., Albajes, R., Lopez, C., Eras, J., Baraibar, B., Lumbieres, B. & Pons, X. 2006: Transgenic Bt maize: main results of a six-year study on non-target effects. IOBC/WPRS Bull. 29(5): 49-55.

Garcia-Alonso, M., Jacobs, E., Raybould, A., Nickson, T.E., Sowig, P., Willekens, H., van der Kouwe, P., Layton, R., Amijee, F., Fuentes, A.M. & Tencalla, F. 2006: A tiered system for assessing the risk of genetically modified plants to non-target organisms. Environ. Biosafety Res. 5: 57-65.

Hill, R.A. & Sendashonga, C. 2003: General principles for risk assessment of living modified organisms: Lessons from chemical risk assessment. Environ. Biosafety Res. 2: 81-88.

Lövei, G.L. & Arpaia, S. 2005. The impact of transgenic plants on natural enemies: a critical review of laboratory studies. Ent. Exp. Appl. 114: 1-14.

Lozzia, G.C. 1999: Biodiversity and structure of ground beetle assemblages (Coleoptera: Carabidae) in Bt corn and its effects on non target insects. Boll. Zool. Agr. Bachic., Ser. II 34: 37-58.

Lozzia, G.C., Furlanis, C., Manachini, B. & Rigamonti, I.E. 1998: Effects of Bt corn on Rhopalosiphum padi L. (Rhynchota: Aphididae) and on predatopr Chrysoperla carnea Stephen (Neuroptera: Chrysopidae). Boll. Zool. Agr. Bachic., Ser. II 30: 153-164.

Lumbierres, B., Albajes, R. & Pons, X. 2004: Transgenic Bt maize and Rhopalosiphum padi (Hom., Aphididae) performance. Ecol. Entomol. 29: 309-317.

Meissle, M., Vojtech, E. & Poppy, G.M. 2004: Implications for the parasitoid Campoletis sonorensis (Hymnenoptera, Ichneumonidae) when developing in Bt maize-fed Spodoptera littoralis larvae (Lepidoptera: Noctuidae). IOBC/WPRS Bull. 27(3): 117-123.

Romeis, J., Meissle, M. & Bigler, F. 2006: Transgenic crops expressing Bacillus thuringrensis toxins and biological control. Nature Biotechnology 24: 63 – 71.

Saxena, D. & Stotzky, G. 2001: Bt corn has a higher lignin content than non-Bt corn. Am. J. Bot. 88: 1704-1706.

Sehnal, F., Habustova, O., Spitzer, L., Hussain, H.M. & Ruzicka, V. 2004: A biannual study on the environment impact of Bt maize. IOBC/WPRS Bull. 27(3): 147-160.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 59-66 Round robin quantitation of Cry3Bb1 using the qualitative PathoScreen ELISA Hang Thu Nguyen1, Heinz Hunfeld2, Michael Meissle3, Rona Miethling-Graff4, Sibylle Pagel-Wieder2, Stefan Rauschen5, Corinne Zurbruegg6, Sabine Eber5, Frank Gessler2, Jörg Romeis3, Christoph C. Tebbe4, Wolfgang Nentwig6, Johannes A. Jehle1 1Agricultural Service Center Palatinate (DLR-Rheinpfalz), Dept. of Phytopathology, Laboratory for Biotechnological Crop Protection, Breitenweg 71, D-67435 Neustadt/Wstr., Germany; 2Institute of Applied Biotechnology in the Tropics at the University of Goettingen, Kellnerweg 6, 37077 Goettingen, Germany; 3Agroscope Reckenholz-Tänikon Research Station ART, Reckenholzstr. 191, 8046 Zurich, Switzerland; 4Federal Agricultural Research Centre (FAL), Institute of Agroecology, Bundesallee 50, D-38116 Braunschweig, Germany; 5RWTH Aachen University, Institute for Environmental Research (Biology V), Worringerweg 1, D-52054 Aachen, Germany; 6University of Bern, Community Ecology, Baltzerstrasse 6, 3012 Bern, Switzerland. (E-mail: hang.nguyenthu@dlr.rlp.de, johannes.jehle@dlr.rlp.de) Abstract: The potential of quantitative detection of Cry3Bb1 protein using a commercially available qualitative enzyme-linked immunosorbent assay (ELISA) kit (PathoScreen) was evaluated based on a round robin test in six different laboratories in Germany and Switzerland. Three standardized sources of purified Cry3Bb1 protein, one sample with a Cry3Bb1 concentration unknown to the experimenters, and two standardized plant samples of the transgenic maize event MON88017 were measured by all laboratories. Different extraction methods and different incubation conditions for the ELISA were used by the different laboratories. The variability of the quantitation of Cry3Bb1 protein was ± 16.9% among assays in the same laboratory. Among different laboratories, the variability was also observed. Our results indicated that not the ELISA conditions but the extraction methods were the major factors contributing to the variation among the laboratories. In addition, different dilution levels of the antigen also contribute to the variability of quantitation of the Cry3Bb1 protein. Key words: ELISA, Cry3Bb1 protein, PathoScreen kit, MON88017 Introduction Currently, different qualitative and quantitative methods, such as polymerase chain reaction (PCR) and enzyme-linked immunosorbent assay (ELISA) are commercially available for detecting genetically modified organisms (GMOs). Many ELISA kits have been commercialized as a useful and quick tool for detecting Cry1Ab, Cry1Ac, Cry3Bb1 and others in GM plants carrying genes of different Bacillus thuringiensis (Bt) strains. These kits allow the detection of Bt proteins in plants, arthropods, and soil samples (Baumgarte & Tebbe, 2005; Nguyen & Jehle, 2007). For the verification of the Bt-Cry3Bb1 protein there are different commercial ELISA kits available, e.g. the PathoScreen kit for Bt-Cry3Bb1 protein (Agdia, Elkhart, IN, USA) or the QualiPlateTM kit for YieldGard®Rootworm corn – AP 015 (Envirologix, Portland, ME, USA). However, these kits are available only for a qualitative detection. Nevertheless, Cry3Bb1 protein in soil was quantitatively determined using the PathoScreen kit (Ahmad et al., 2005, 2006). In order to test the ability of this kit for quantifying Cry3Bb1 protein, and to investigate the reproducibility and robustness of its quantification under different laboratory conditions, a round robin test in six different

59

60

laboratories in Germany and Switzerland was performed. The tests were carried out using three standardized sources of the purified Cry3Bb1 protein, two standardized Bt-maize samples of the event MON88017, and one non Bt-maize sample. The data obtained in the different laboratories are compared. Materials and Methods Cry3Bb1 protein standards and standardized test material Three different Cry3Bb1 protein standards were used: Standard A: approx. 40 ng ml-1 of lyophilised protein, delivered with the PathoScreen kit

(Agdia). The concentration was given by the supplier. Standard M: 10 µg ml-1 of purified solubilized Cry3Bb1 protein expressed in Escherichia coli

using the expression plasmid pMON70855 (provided by Monsanto company). The concentration was evaluated from amino acid composition data.

Standard N: 10 µg ml-1 of protein expressed and purified at the DLR Rheinpfalz using the plasmid pMON70855 (plasmid provided by Monsanto company). The concentration was determined according to Bradford (1976).

The following test samples were used: Blind sample B: purified Cry3Bb1 protein sample derived from standard N with a

concentration of 50 µg ml-1, but not known to the different experimenters. Sample MI: finely chopped maize leaves from event MON88017, frozen. Sample MII: lyophilized leaf powder from event MON88017. Sample MIII: finely chopped conventional maize leaves from the variety Benicia, frozen. The maize samples were provided by the Institute for Environmental Research at RWTH Aachen University.

Experimental design The commercially available double antibody sandwich ELISA “PathoScreen kit for Bt-Cry3Bb1 protein” (Agdia, Elkhart IN, USA) was used to determine the Cry3Bb1 protein concentrations in standards A, M, N, in the blind sample and in the different maize leaf samples. The round robin measurements were independently performed on different days in six laboratories. Seven-point standard curves with concentrations of 0.59, 0.89, 1.33, 2.00, 3.00, 4.50 and 6.75 ng ml-1 were established for each of the protein standards A, M and N. Only standard A of laboratory 5 was diluted to 5 different concentrations 0.625, 1.25, 2.5, 5.0, 10 ng ml-1. In order to quantify the Cry3Bb1 protein content in the samples the standard curve of standard A was used.

The Cry3Bb1 protein from the maize leaf samples was extracted using different methods established in the different laboratories (Table 1). The resulting homogenates were centrifuged (Table 1) and the supernatants were diluted in phosphate buffered saline Tween 20 (PBST) supplied with the PathoScreen kit. Each sample was tested in duplicate or triplicate. Except for the laboratory specific ELISA conditions (incubation times and temperatures, see Table 1) the test was performed according to the manufacturer’s instructions. Briefly, the sample solutions were pipetted into test wells, which were coated with polyclonal antibodies against Cry3Bb1 protein. Cry3Bb1 protein captured by the antibodies was detected by a peroxidase enzyme-conjugated secondary antibody. After washing with PBST buffer, a 3,3',5,5'-tetramethylbenzidine (TMB) substrate was added to the plate for a colour development step to visualise the results of the assay. Laboratory 5 measured the optical density (OD) at 630 nm. All other laboratories measured at 450 nm after

61

stopping the colour development with 3M sulphuric acid. The lot numbers of the different ELISA kits and the corresponding buffers are listed in Table 2.

Table 1. Extraction methods and incubation conditions used in the participating laboratories for the leaf sample analyses.

Incubation Laboratory Equipment and extraction methods

Centrifugation

time (h) temperature (°C)

1 Ultra-Turrax homogenizer (Kinematica, Switzerland) (30, 000 rev/min for 1 min)

5,000 xg, 5 min 15 4

2 Retsch bead mill (30 Hertz for 3 min 40 sec )

12,000 xg, 5 min 2 20-25

3 liquid nitrogen, mini pestle, vortex mixer

10,000 xg, 5 min 2 20-25

4 liquid nitrogen, mini pestle 12,000 xg, 5 min 16 4 5 Bioreba homogenizer and

extraction bags 600 xg, 10 min 16 4

6 PBST buffer, overhead shaker (15 min), no mechanical mazeration

23,000 xg, 5 min 18 4

Table 2. Lot numbers of the kit components applied in the different laboratories. Laboratory Antibody-coated

96-well microtiter plate

Enzyme conjugate

TMB substrate

Positive control

PBST buffer Sulphuric acid

1 00033 00089 00319 C1446 00310 AppliChem 2 00036 00083 00331 C1446 00310 not specified 3 00033 00097 00319 C1446 00310 0089 (Agdia) 4 00033 00079 00319 C1205 00315 Agdia 5 00036 00083 00331 C1133 00077 - 6 00033 00049 00310 C1446 00310 Merck

Results and Discussion Comparison of the 3 standard curves In the ELISAs measured at 450 nm the maximum OD values reached 0.5 to 1.2 when the samples were incubated at room temperature for 2 h (Figure 1-2; 1-3). When the incubation time was extended to 15-18 h at 4°C, the OD values increased to 0.7-3.1 at 450 nm (Figure 1-1; 1-4; 1-6), while at 630 nm the OD values were only 0.4-1.3 (Figure 1-5). These results show that all different incubation conditions (room temperature for 2 h vs. overnight at 4°C) resulted in a linear protein concentration line. They caused differences only in the absolute height of their OD values; extended incubation at lower temperature increased the sensitivity of the test by a factor of three (Figure 1). The nominal concentrations of the three protein standards were identical, which suggested that the standard curves should be similar. However, in most laboratories standard A showed slightly higher ODs than the standards N

62

and M (Figure 1). This could be explained by the fact that the concentration of standard A was probably higher than those of M and N, the concentrations of which were determined using the amino acid composition and the Bradford method, respectively.

A (R2 = 0.9947)M (R2 = 0.9991)N (R2 = 0.9906)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9845) M (R2 = 0.986) N (R2 = 0.9765)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9813)M (R2 = 0.9956)N (R2 = 0.9925)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9976)M (R2 = 0.9961)N (R2 = 0.9958)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9945)M (R2 = 0.9958N (R2 = 0.9914)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9593)M (R2 = 0.9983)N (R2 = 0.9984)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

Cry3Bb1 concentration [ng/ml] Cry3Bb1 concentration [ng/ml]

A M N

OD

450

nm

OD

630

nm

OD

450

nm

OD

450

nm

OD

450

nm

OD

450

nm

1

2

3

4

5

6

A (R2 = 0.9947)M (R2 = 0.9991)N (R2 = 0.9906)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9845) M (R2 = 0.986) N (R2 = 0.9765)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9813)M (R2 = 0.9956)N (R2 = 0.9925)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9976)M (R2 = 0.9961)N (R2 = 0.9958)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9945)M (R2 = 0.9958N (R2 = 0.9914)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9593)M (R2 = 0.9983)N (R2 = 0.9984)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

Cry3Bb1 concentration [ng/ml] Cry3Bb1 concentration [ng/ml]

A M N

OD

450

nm

OD

630

nm

OD

450

nm

OD

450

nm

OD

450

nm

OD

450

nm

A (R2 = 0.9947)M (R2 = 0.9991)N (R2 = 0.9906)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9845) M (R2 = 0.986) N (R2 = 0.9765)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9813)M (R2 = 0.9956)N (R2 = 0.9925)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9976)M (R2 = 0.9961)N (R2 = 0.9958)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9945)M (R2 = 0.9958N (R2 = 0.9914)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

A (R2 = 0.9593)M (R2 = 0.9983)N (R2 = 0.9984)

0,00,51,01,52,02,53,03,5

0 2 4 6 8 10 12

Cry3Bb1 concentration [ng/ml] Cry3Bb1 concentration [ng/ml]

A M NA M N

OD

450

nm

OD

630

nm

OD

450

nm

OD

450

nm

OD

450

nm

OD

450

nm

1

2

3

4

5

6

Figure 1. Standard curves of the different Cry3Bb1 protein standards A (■), M (●), and N (▲) established by the different laboratories (1-6). [Lab.1: A: y = 0.4270x – 0.0124; M: y = 0.4650x – 0.0552; N: y = 0.3671x – 0.0892 ; Lab.2: A: y = 0.1288x – 0.0587; M: y = 0.0941x – 0.0461; N: y = 0.0733x – 0.0472 ; Lab.3: A: y = 0.1731x – 0.0258; M: y = 0.1258x – 0.0034; N: y = 0.1345x – 0.0526 ; Lab.4: A: y = 0.4356x – 0.0645; M: y = 0.1090x – 0.0370; N: y = 0.2935x – 0.1123 ; Lab.5: A: y = 0.1313x + 0.0066; M: y = 0.0533x – 0.0149; N: y = 0.0764x – 0.0218 ; Lab.6: A: y = 0.4440x + 0.1215; M: y = 0.4412x – 0.1191; N: y = 0.4090x -0.1223].

Three independent measurements were performed in laboratory 1 using protein standard A and applying the same conditions. The repeatability (r) ranged from 7.5 to 16.9% RSDr (relative standard deviation). Hence, the precision of quantification using the PathoScreen kit may be lower than 16.9%.

63

Quantitation of purified Cry3Bb1 Cry3Bb1 protein of the blind sample B was quantified by six laboratories, in order to examine the variation of quantitation. Four laboratories (1, 2, 4, 6) determined the Cry3Bb1 concentration in the range of 25-34 µg ml-1, whereas two laboratories (3, 5) failed to quantify the protein sample due to improper dilutions (Table 3). The average recovery was between 50 and 68% based on standard A, which was supplied with the kit. As noted above, standard A gave generally higher OD values than standard N. Hence, the protein concentration of the sample was systematically underestimated by standard A. Variation was observed among the laboratories as well as among the different dilutions used in the ELISA detection. The variation coefficients of the dilutions (CVr) ranged from 2.5 to 16.6% and the reproducibility coefficient of variation (CVR) among the laboratories was 15%. Thus, the dilution-dependent variability of the ELISA can be as big as the inter-laboratory variability of the measurements, underlining the importance of making appropriate dilutions for the ELISA. Table 3. Quantitation of Cry3Bb1 protein in the blind sample B and precision parameters. Mean of the Cry3Bb1 protein, standard deviation and coefficient of variation were calculated for different dilutions in each laboratory and for all the dilutions and laboratories (except for laboratory 3 and 5).

Laboratory

Dilution Cry3Bb1 (µg ml-1)

Mean of different dilutions

Sr1

CVr2 (%)

among different dilutions

1 1:10,000 33.0 34.0 1.5 4.5 1:20,000 35.1

2 1:10,000 28.2 25.2 4.2 16.6 1:20,000 22.3

3 1:1 0.014 n.d. n.d n.d. 4 1:5,000 31.0 30.5 0.8 2.5 1:10,000 29.9 5 1:250 7.10 n.d. n.d. n.d. 6 1:12,500 28.0 25.2 3.5 14.1 1:20,000 26.4 1:25,000 21.2

Mean of laboratories 1,2,4,6 28.7 SR

3 4.3 CVR 4(%) 15.0

1repeatability standard deviation; 2repeatability coefficient of variation; 3reproducibility standard deviation, 4reproducibility coefficient of variation; n.d. not determined due to improper dilutions.

64

Quantitation of Cry3Bb1 in standardized Bt-maize samples Mon88017 Quantitation of Cry3Bb1 protein in two standardized Bt-maize samples MI and MII were performed using the laboratory specific extraction methods and ELISA incubation parameters as given in Table 1. Different dilutions were applied in the tests and only those OD values, which lay within the standard curve of the PathoScreen kits were considered for evaluation (Table 4 and Table 5). Laboratory 6 used a rather inefficient protein extraction method and the quantified Cry3Bb1 content was very low. Hence, the results of this laboratory were excluded from further statistical analysis. For MI the Cry3Bb1 measurements ranged from 42.3 to 135.0 µg g-1 fresh weight (fw), with a mean of 92.4 µg g-1 fw. The quantified Cry3Bb1 contents varied depending on the laboratories and the antigen dilutions applied in the ELISA measurements. The CVr ranged from 3.1 to 30.1% (Table 4). The CVR was 36.9% among five laboratories using different extraction methods and dilutions (Table 4). Table 4. Detection of Cry3Bb1 protein in MI using PathoScreen Kit. Mean of the Cry3Bb1 protein, standard deviation and coefficient of variation were calculated for different dilutions in each laboratory and for all the dilutions and laboratories (except for laboratory 6).

Laboratory Dilution Cry3Bb1 (µg g-1 fw)

Mean of dilutions

Sr1 CVr

2 (%)

1 1:24,000 87.5 87.5 - - 2 1:50,000 86.4 88.4 2.7 3.1 1:50,000 90.4

3 1:5,000 40.7 42.3 2.2 5.2 1:10,000 43.8

4 1:50,000 85.8 109.0 32.8 30.1 1:100,000 132.2

5 1:19,000 135.0 135.0 - - 6 1:10 0.059 n. d. n.d. n.d.

Mean of laboratories 1,2,3,4,5 92.4 SR3 34.1

CVR4 (%) 36.9

1repeatability standard deviation; 2repeatability coefficient of variation; 3reproducibility standard deviation, 4reproducibility coefficient of variation; n.d. not determined due to improper dilutions.

For MII the Cry3Bb1 measurements ranged from 23.9 to 57.2 µg g-1 dry weight (dw); the mean was 36.2 µg g-1 dw (Table 5). The CVr ranged from 4.5 to 25.8% among dilutions from each laboratory (Table 5), with a CVR of 36.2% among five laboratories.

There was a clear correlation between the measurements of MI and MII (r2 = 0.88). Those laboratories, which measured the highest Cry3Bb1 contents in MI, also did for MII, and those, which measured the lowest contents, also did for both. Therefore, it is obvious that the measurement of Cry3Bb1 in maize tissues strongly depends on the extraction efficacy. Limit of detection The limit of detection (mean + 3xSD of the non Bt-maize sample MIII) was 0.5 ng g-1 and 0.3 ng g-1, which was determined by the two laboratories 1 and 5, respectively.

65

Table 5. Detection of Cry3Bb1 protein in MII using PathoScreen Kit. Mean of the Cry3Bb1 protein, standard deviation and coefficient of variation were calculated for different dilutions in each laboratory and for all the dilutions and laboratories (except for laboratory 6).

Laboratory Dilution Cry3Bb1 (µg g-1 dw)

Mean of dilutions

Sr1 CVr

2 (%)

1 1:6,000 33.9 36.4 2.4 6.5 1:12,000 36.8 1:20,000 38.6

2 1:50,000 28.0 26.4 2.3 8.7 1:50,000 24.8

3 1:1,000 24.9 23.9 1.1 4.5 1:5,000 22.8 1:10,000 24.1

4 1:10,000 43.9 37.2 9.6 25.8 1:50,000 30.4

5 1:12,000 52.2 57.2 7.0 12.2 1:5,000 62.1

6 1:10 0.063 n.d. n.d. n.d. Mean of laboratories 1,2,3,4,5 36.2 SR

3 13.1 CVR

4(%) 36.2 1repeatability standard deviation; 2repeatability coefficient of variation; 3reproducibility standard deviation, 4reproducibility coefficient of variation; n.d. not determined due to improper dilutions. Conclusions The comparison of six laboratory procedures showed that the PathoScreen kit is potentially useful for the quantitation of Cry3Bb1 protein. The variability among the laboratories was mainly derived from (i) the Cry3Bb1 extraction method, (ii) the antigen dilutions used in the ELISA, and (iii) the source of the Cry3Bb1 standard. Further optimisation and standardisation may result in a validated test that can be used for a reliable quantitation of Cry3Bb1 in transgenic plant material. Acknowledgements We thank Monsanto company for providing the Cry3Bb1 protein, the plasmid pMON78055, and seeds of the maize event MON88017 and control line. The four participating laboratories in Germany were supported by grants of the Federal Ministry of Education and Research (BMBF) within the research consortium “Post-market safety research on transgenic maize with new Bt genes”. The laboratories in Switzerland were supported by grants from the Federal Office for the Environment (FOEN) and the Swiss Innovation Promotion Agency (CTI). References Ahmad, A, Wilde, G.E. & Yan Zhu, K. 2005: Detectability of Coleopteran-specific Cry3Bb1

protein in soil and its effect on nontarget surface and below-ground Arthropods. Environ. Entomol. 34: 385-394.

66

Ahmad, A., Wilde, G.E., & Yan Zhu, K. 2006: Evaluation of effects of Coleopteran-specific Cry3Bb1 protein on earthworms exposed to soil containing corn roots or biomass. Environ. Entomol. 35: 976-985.

Baumgarte, S. & Tebbe, C.C. 2005: Field studies on the environmental fate of the Cry1Ab Bt toxin produced by transgenic maize (MON810) and its effect on bacterial communities in the maize rhizosphere. Mol. Ecol.14: 2539−2551.

Bradford, M. 1976: A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248-254.

Nguyen, H.T. & Jehle, J.A. 2007: Quantitative analysis of the seasonal and tissue-specific expression of Cry1Ab in transgenic maize MON810. J. Plant Dis. Prot. 114: 82-87.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 67-74 F2 Screen and field sampling with light trap cages, two methods for a resistance monitoring in Bt crops Heike Engels, Ingolf Schuphan, Sabine Eber Institute for Environmental Research (Biologie V), Aachen University, Aachen, Germany (E-mail: engels@bio5.rwth-aachen.de)

Abstract: Large-scale cultivation of Bt crops will exert high selection pressure on the target pest, which may consequently evolve resistance. So far, no resistance of the European corn borer Ostrinia nubilalis to Bt crops has been reported. As yet, no anticipatory resistance monitoring plan has been established for Europe.

When resistance alleles are rare, the most efficient method for a resistance monitoring is assumed to be the F2 screen. This method preserves the genetic variation among isofemale lines and concentrates potential resistance alleles into homozygous genotypes of the F2 generation. This way it is possible to estimate the frequency of resistance alleles in field populations. In our study 450 isofemale ECB lines, started from 650 females from field populations in four German regions, were screened over three cultivation periods.

Simultaneously a simpler monitoring method has been developed and tested. Target pest insects are thereby attracted to light-trap cages containing insect resistant transgenic plants. Trapped target insects lay their eggs onto these plants and hatching neonates will thus be screened for resistance by feeding on them. For resistance estimates the number of females and egg masses can be quantified in the light-trap cages throughout the season. In a single test trial during one cultivation period in one cage 1670 egg masses and approx. 50,000 neonate larvae were screened.

The two methods are compared in terms of their expenses, requirements and suitability for a long-term resistance monitoring during the cultivation of transgenic crops. Whilst the F2 screen enables to detect resistance development at an early state at low resistance allele frequencies, it is also a very expensive method, both in terms of finances and time. Monitoring by light trap cages, on the other hand, is a less sensitive method, as it can only detect resistance at the homozygous stage. It is, however, a very simple and unexpensive method, which makes it possible to screen large numbers of target pest insects in the field . Key words: Bt crops, resistance management, target pest, lepidopterans, monitoring method, Ostrinia nubilalis Introduction Insecticidal transgenic crops using Bt technology, are widely planted in the USA, Argentina, Brazil, Canada, China and Spain. Bt maize provides control of lepidopteran maize pests without the use of conventional insecticides. However, large scale cultivation of Bt crops will exert high selection pressure on the target pest species, which may consequently evolve resistance and thus eliminate the benefits of these crops.

Several techniques have been used to estimate the frequency of rare resistance alleles in natural populations. When resistance alleles are believed to be rare, Andow & Alstad (1999) and Andow & Ives (2002) proclaim that the most efficient method is an F2 screen, which has been introduced by Andow & Alstad (1998). The F2 screen preserves genetic variation among isofemale lines and concentrates resistance alleles into homozygous genotypes in the F2 generation. So far, F2 screens have been used for the European corn borer (ECB, Ostrinia nubilalis) (Andow et al., 1998; Andow et al., 2000; Bourguet et al., 2003; Stodola et al.,

67

68

2006), diamondback moth (Plutella xylostella) (Zhao et al., 2002), rice stem borer (Scirophaga incertulas) (Bentur et al., 2000), cotton bollworm (Helicoverpa armigera) (Akhurst et al., 2003), cottonwood leaf beetle (Chrysomela tremulae) (Génissel et al., 2003), and sugarcane borer (Diatraea saccharalis) (Huang et al. submitted). No resistance alleles have been detected in populations of ECB or the rice stem borer, but alleles have been found in F2 screens of natural populations of the cotton bollworm, cottonwood leaf beetle and the sugarcane borer. However, even for species where alleles have not been found, the F2 screen has allowed an estimation of the frequency of resistance alleles in natural populations, which has been valuable for resistance management planning.

More recently a simple long-term monitoring method has been developed and tested. This method is described below. It combines mass capturing of adult pest insects with an autocidal procedure for the selection of resistant individuals from field populations. We performed and compared both methods in terms of their expenses, requirements and suitability for a long-term resistance monitoring during the cultivation of transgenic crops. Material and Methods F2 screen The F2 screen protocol for O. nubilalis (Lepidoptera: Crambidae) is described in Andow & Alstad (1998). Mated females are considered to be the preferred stage for initiating an F2 screen, but many variant methods have been proposed (Bentur et al., 2000, Zhao et al., 2002, Stodola & Andow, 2004, Stodola et al., 2006). For our F2 screens either mated females, pair-mated adults of late instar larvae, or egg masses, all from field populations, were used to establish isofemale lines. From each isofemale line the F1 offspring was reared and sib-mated, and finally the F2 neonates were screened on Bt plants or Bt diet. Starting from the two P1 adults, each isofemale line allowed us to characterise at least four haplotypes. By sib-mating the F1 generation, 1/16 of the F2 individuals are expected to be homozygous for the resistance allele if one of the P1 adults carried a resistance allele. (i) Insect collection and rearing ECB sampling took place in 2003-2005 at four different locations in Germany (Table 1). During summer adults were collected using light traps. Moths were transferred in pairs to small cages (10 cm Ø x 10 cm) covered with waxed paper for oviposition. The purpose of caging females with males was to ensure the mating of females that might not have mated in the field.

Late instar larvae of O. nubilalis were collected during fall 2003 and 2004 from Bonn and Heilbronn. Thereby only one larva was taken per maize plant to minimise the probability that siblings were collected. Larvae were kept in groups of 15 individuals in petri dishes and provided with a special ECB diet. They were kept at 25oC, 70 % RH and continuous day light to avoid diapause. Pupation time varied strongly between individuals and many larvae did not pupate at all. Once a week the dishes were checked for pupae, which were removed and males and females were kept separately in plastic boxes until the adults emerged. After emergence, males and females were paired in similar cages as the field-collected individuals. Moths were provided with a solution of 10% honey in water and stored at 25oC, 70 % RH and a 16:8 h (L:D) photoperiod. (ii) F1 rearing and F2 screening After mating the P1 adults in the cages, egg masses were daily collected, placed in petri dishes with ECB diet and stored at 25oC, 70 % RH and a 16:8 h (L:D) photoperiod until the blackened head capsules were visible within the egg. To synchronise the offspring of one

69

female, egg masses were then kept at 10°C until the female had finished oviposition. After that all life stages of the offspring were reared at 25oC, 70 % RH and a 16:8 h (L:D) photoperiod. Larvae were weekly transferred to new dishes containing fresh diet, and after the 2nd larval instar their number was reduced to 15 individuals per dish. F1 pupae from different families were kept seperately, and emerged adults were daily transferred to F1 family cages for sib-mating. Moths were provided with a solution of 10% honey in water to maximise the oviposition rate.

Neonates hatched from F2 egg masses were screened on Bt maize leaves (Novelis, event Mon 810, Monsanto Company). F2 screens were performed in plastic boxes (10 cm Ø x 10 cm). Thereby five to ten ECB egg masses were transferred into each box together with 6-8 leaves of Bt-maize tissue over a moistened filter paper. Every third day new Bt leaves were added and the filter paper was remoistened. The boxes were stored at the same conditions as the F1 generation. All dishes were checked for surviving larvae after 7 to 10 days. Resistance was only assumed when larvae survived beyond the 2nd larval instar with additionally evidence of feeding on Bt maize leaves. Field sampling with light trap cages The tested light trap method combines mass capturing of adult pest insects with an autocidal procedure for the selection of resistant individuals from field populations. The light-trap consists of a lamp connected to a cage (Figure 1). At the upper lamp funnel a hook is attached by which the lamp is fixed to a crossbar, which carries the electric cable. The lamp is constructed of two plastic funnels (Ø 25 cm) between which two brackets hold a central fluorescent tube (15 W/05 Philips TLD). Four vertical acrylic glass wings (60 x 10 x 0.2 cm, Figure 1) are arranged around the light tube at right angles to each other. The cage frame is constructed from alloy tubing (200 x 150 x 200 cm) covered by a commercially available mosquito net. Within the cage appropriate transgenic insect-resistant plants are placed.

By switching on the light at suitable times the target pest insects are attracted, collide with the glass wings and fall through the lower funnel into the cage. Trapped females lay their eggs onto the insecticidal plants and hatching neonates feed on them. This way neonates are screened for susceptibility, so that only resistant larvae will survive and develop on the plants. At the end of the cultivation period surviving larvae can be recovered and used for further studies on pest resistance. As a basis for estimates of resistant allele frequencies the number of captured females in the cage and the number of egg masses on the transgenic plants can be quantified during the cultivation period. Test trial During a first experiment using ECB as the target pest, a light-trap cage following the above description was set up in Southern Germany in a conventional maize field. Within the cage, sixteen Bt-maize plants (Novelis event Mon 810) were planted on June 21, 2005. Starting from this date until September 22, 2005, the lamp was switched on each night from 21:30 to 24:00 h. The light-trap cage was checked at the dates listed in Table 2. During each survey all ECB adults were counted and all new egg clutches on the Bt plants were quantified and marked, but left on the plants. Older egg clutches were checked for surviving neonate larvae and all plants were carefully examined for signs of ECB feeding. After the ECB flight period, the maize plants were harvested and dissected for feeding damage and surviving ECB larvae.

70

Bt-maize

Figure 1. Light trap consisting of gallows, fluorescent lamp and cage and transgenic maize plants. Results F2 screen In total 455 isofemale lines and 225,615 individuals from the four German locations were screened (Table 1). There were no surviving F2 larvae, hence no resistant alleles were present in these populations. Consequently the expected frequencies of resistance alleles in Germany are very low.

As general described in this procedure overall losses during the F2 screen were quite high (Table 1), as moths died before oviposition or did not mate, egg masses did not hatch or larvae did not pass diapause (Andreadis et al., 2007; Bourguet et al., 2003). As an estimate of the expenses for the F2 screen we arrived at costs of € 500 per screened line, which includes scientific personnel for insect collection and rearing, travel costs for field sampling, and all expenses for consumables and associated infrastructure. If experienced non-scientific staff is available to perform this method, expenses could be reduced.

71

Table 1. Number of lines progressing through the F2 screen experiment

Year Location P lines started F1 lines Screened F2 lines

2003 Bonn 37 15 9 2004 Heilbronn 36 21 21

Heilbronn– Diapaused ~150 100 93

Würzburg 82 39 39

2005 Walldorf - Egg masses 200 167 153

Walldorf - Paired P 144 144 140

Total 657 494 455

Field screening with light trap cages In total, 270 ECB females and 1667 egg masses were found in the cages and on the encaged Bt-plants. Each egg clutch contained about 10-50 eggs from which the neonates hatched. Hence approximately 50,000 neonate larvae were screened (Table 2), none of which survived on the Bt-plants. After harvest (September 22, 2005) the 16 Bt-corn plants were dissected for the presence of O. nubilalis larvae. No surviving larvae were found within the plants. In order to compare both methods, procedure costs for the field sampling method were also estimated. The light trap cage construction caused costs of approximately € 100. Travel and personnel costs are dependent on how the monitoring is performed. Assuming that local farmers take part in the implementation of this method, total average costs of € 200 were estimated per cage and cultivation period. Table 2. Number of moths caught in the trap, number of new egg masses and estimated number of eggs per date Date Number of moths Number of new egg

masses Estimated number

of eggs 21 June 2005 Bt-maize planted 25 June 2005 50 53 1590 28 June 2005 100 298 8940 06 July 2005 50 984 29520 12 July 2005 50 198 5940 19 July 2005 20 181 5430 28 July 2005 0 3 90 10 August 2005 0 0 0 22 September 2005 Harvest Total 270 1,667 50,010

72

Discussion In terms of a long-term monitoring of target pest resistance to transgenic crops both methods discussed here have particular strengths as well as weaknesses. The major advantage of the F2 screen is its sensitivity. It detects resistance development at an early stage at very low resistance allele frequencies, because it gives evidence of recessive resistance alleles in heterozygous individuals. On the other hand this method is rather time consuming, complex and thus expensive considering a long-term application. It requires rearing the target insects over several generations which may pose practical problems, e.g. at preventing diapause, and makes it quite a protracted method, dependent on the particular species. Due to its sophisticated procedure it also needs to be performed by skilled staff experienced in insect rearing und susceptibility testing.

Monitoring by light trap cages, on the other hand, is a considerably less sensitive method, as it can only detect resistance at the homozygous stage. This means that the frequency of a recessive resistance allele may be comparably high, when resistance development is detected. However, this method it is simple to apply, requires low maintenance and costs, and therefore could be performed by local farmers.

Light-trap cages have been used in our research group to study European corn borer moths for more than a decade (Saeglitz, 2003; Saeglitz et al., 2006). The procedure is generally suitable for nocturnal Lepidopterans and may most likely also be suitable for other pests in maize. From experience we know that the capture rate depends on the pest density near the light-trap cage. Hence it is possible to place light-trap cages in locations where a high pest density is expected. Screening for resistance by using this autocidal method can be performed in Bt-crop fields as well as in non-Bt fields, e.g. in refuges. Two options exist: to run the traps without weekly control until harvest or to count captured adults and mark all egg masses on a regular basis. In the latter case the occurrence of surviving larvae will give an early warning of resistance and can be followed by immediate resistance management measures. The total number of captured females and egg masses will then provide the basis for a statistical evaluation of resistance allele frequency.

The screening results of both monitoring methods lead to a comparison of the expenses of those methods. We estimated approximate costs of € 500 per line for the F2 screen, which results in total costs of € 227,500 for 455 isofemale lines and 225,615 screened individuals. These expenses are accordant to those of other experts in F2 screen processing in Spain and Greek (Andreadis et al., 2007). The estimated costs per light trap per cultivation period was € 200. Hence using the expenses for the above F2 screen it would be possible to set up 1137 light-trap cages in the field. This way ~ 57 million pest insect individuals could be screened with this field sampling method, assuming a similar efficiency as in out test trial. From a practical point of view it seems reasonable to use both methods, but on different temporal and spatial scales. The F2 screen should be used as a random initial test to get an idea of resistance allele frequency in field populations, whereas field sampling with light-trap cages should be used for routine monitoring at a much wider spatial scale. Due to the large number of screened target insects using this cost efficient method in many regions may considerable increase the possibility of resistance detection. Acknowledgements We thank Arti Sinha, Ulrike Schuller, Katharina Pietsch, Kai Priesnitz, Claudia Gaspers and Stefan Rauschen for helping to assemble the light trap cages and for assistance during the field and laboratory work. We also thank Johannes Lemburg for drawing the light-trap cage construction.

73

References Akhurst, R.J., James, W., Bird, L.J. & Beard, C. 2003: Resistance to the Cry1Ac delta-

endotoxin of Bacillus thuringiensis in the cotton bollworm, Helicoverpa armigera (Lepidoptera: Noctuidae). J. Econ. Enomol. 96: 1290-1299.

Andow, D.A. & Alstad, D.N. 1998: F2 Screen for rare resistance alleles. J. Econ. Entomol. 91: 572-578.

Andow, D.A. & Alstad, D.N. 1999: Credibility interval for rare resistance allele frequencies. J. Econ. Entomol. 94: 755-758.

Andow, D.A., Alstad, D.N., Pang, Y.-H., Bolin, P.C. & Hutchison, W.D. 1998: Using an F2 screen to search for resistance alleles to Bacillus thuringiensis toxin in European corn borer (Lepidoptera: Crambidae). J. Econ. Entomol. 91: 579-584.

Andow, D.A. & Ives, A.R. 2002: Monitoring and adaptive resistance management. Ecol. Appl. 12: 1378-1390.

Andow, D.A., Olson, D.M., Hellmich, R.L., Alstad, D.N. & Hutchison, W.D. 2000: Frequency of resistance to Bacillus thuringiensis toxin Cry1Ab in an Iowa population of European corn borer (Lepidoptera: Crambidae). J. Econ. Entomol. 93: 26-30.

Andreadis, S.S., Alvarez-Alfageme, F., Sanchez-Ramos, I., Stodola, T.J., Andow, D.A., Milonas, P.G., Savopoulou-Soultani, M. & Castanera, P. 2007: Frequency of Resistance to Bacillus thuringiensis Toxin Cry1Ab in Greek and Spanish Population of Sesamia nonagrioides (Lepidoptera: Noctuidae). J. Econ. Entomol. 100: 195-201.

Bentur, J.S., Andow, D.A., Cohen, M.B., Romena, A.M. & Gould, F. 2000: Frequency of alleles conferring resistance to a Bacillus thuringiensis toxin in a Philippine population of Scirpophaga incertulas (Lepidoptera: Pyralidae). J. Econ. Entomol. 93: 1515-1521.

Bourguet, D., Chaufaux, J., Seguin, M., Buisson, C., Hinton, J.L., Stodola, T.J., Porter, P., Cronholm, G., Buschman, L.L. & Andow, D.A.. 2003: Frequency of alleles conferring resistance to Bt maize in French and US corn belt populations of Ostrinia nubilalis. Theor. Appl. Genet. 106: 1225-1233.

Génissel, A., Augustin, S., Courtin, C., Pilate, G., Lorme, P. & Bourguet, D. 2003 : Initial frequency of alleles conferring resistance to Bacillus thuringiensis poplar in a field population of Chrysomela tremulae. Proc. Roy. Soc. Biol. Sci. Ser. B. 270: 791-797.

Huang, F., Leonard, B.R. & Andow, D.A. 2007: Sugarcane borer (Lepidoptera: Crambidae) resistance to transgenic Bacillus thuringiensis maize. J. Econ. Entomol. 100: 164-171.

Saeglitz, C. 2003: Genetic diversity of European corn borer populations (Ostrinia nubilalis, Hbn.) and their susceptibility for Bacillus thuringiensis Berliner (Bt) - toxin as a basis for resistance management in Bt-corn. Ph.D. dissertation, Aachen University, Aachen, Germany. Available under: http://www.bio5.rwth-aachen.de/oekologie/Mitarbeiter/saeglitz.html http://sylvester.bth.rwth-aachen.de/dissertationen/2004/110/04_110.pdf http://sylvester.bth.rwth-aachen.de/dissertationen/2004/110/04_110_enzu.htm

Saeglitz, C., Bartsch, D., Eber, S., Gathmann, A., Priesnitz, K.U. & Schuphan, I. 2006: Monitoring the Cry1Ab susceptibility of European corn borer (Ostrinia nubilalis Hbn.) in Germany. J. Econ. Entomol. 99: 1768-1773.

Stodola, T.J. & Andow, D.A. 2004: F2 screen variations and associated statistics. J. Econ. Entomol. 97: 1756-1764.

Stodola, T.J., Andow, D.A., Hyden, A.R., Hinton, J.L., Roark, J.J., Buschman, L.L., Porter, P. & Cronholm, G.B. 2006: Frequency of resistance to Bacillus thuringiensis toxin Cry1Ab in southern U.S. Corn Belt population of European corn borer (Lepidoptera: Crambidae). J. Econ. Entomol. 99:502-507.

74

Zhao, J.Z., Y.X. Li, H.L. Collins and A.M. Shelton. 2002. Examination of the F2 screen for rare resistance alleles to Bacillus thuringiensis toxins in the diamondback moth (Lepidoperta: Plutellidae). J. Econ. Entomol. 95: 14-21.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 75-78 Diversity and seasonal phenology of spiders, ground beetles and rove beetles in conventional and transgenic maize in Central Spain Gema P. Farinós1, Marta de la Poza1, 2, Pedro Hernández-Crespo1, Félix Ortego1, Pedro Castañera1* 1 Centro de Investigaciones Biológicas, CSIC, Ramiro de Maeztu 9, 28040 Madrid, Spain (E-mail: castan@cib.csic.es); 2Current address: Centro Nacional de Biotecnología, CSIC, Cantoblanco, 28049 Madrid, Spain Abstract: A farm scale trial was performed during three years to compare the phenology and community structure of the three prevalent groups of aboveground arthropods (spiders, ground beetles and rove beetles) that inhabit Bt maize in Central Spain with those of conventional maize, with and without imidacloprid insecticide seed-treatment. The variability in their activity-density patterns was mostly affected by the year, but no detrimental effects could be associated to transgenic plants. No shifts in richness and diversity indices of spiders, ground beetles and rove beetles were found between conventional and transgenic maize crops, but rove beetles richness was reduced in plots with imidacloprid. Key words: Bt maize, phenology, diversity, non-target predators. Introduction Transgenic maize expressing Cry1Ab toxin from Bacillus thuringiensis (Bt) targeted to control Mediterranean corn borer, Sesamia nonagrioides (Lefebvre) (Lepidoptera: Noctuidae), and European corn borer, Ostrinia nubilalis (Hübner) (Lepidoptera: Crambidae) was approved in Spain in 1998. The first commercialised event was the Bt 176 (var. Compa CB, Syngenta), being practically the only variety grown until 2003 in about 5% of the total maize growing area. Thereafter, the cultivated area of this variety was declining and in 2005 the Spanish Ministry of Agriculture disallow further planting of all Bt 176 varieties. Bt maize varieties derived from the event MON810 were approved in 2003 and reached about 15% of the total maize area in 2006. The use of Bt maize can be considered environmentally friendly due to the specificity of the Cry1Ab toxin towards lepidopteran pests and the reduction of insecticide applications to control corn borers. However, results obtained in some field and laboratory studies warned about potential adverse direct and indirect effects on non-target organisms due to exposure to the insecticidal toxin (Hilbeck et al., 1998; Harwood et al., 2005; Pilcher et al., 2005).

Post-market environmental monitoring plans are only required in the EU since Directive 2001/18/EC, but they are contemplated in the Spanish legislation for the registration of commercial varieties since 1998. As part of this Spanish post-market monitoring plan, a farm scale trial was performed to assess the effect of Compa CB (Event 176) on the abundance and activity-density patterns of predatory arthropods in two Spanish locations from 2000 to 2002 (De la Poza et al., 2005; Farinós et al., 2008).

We report here on the effects of Bt maize on diversity and seasonal phenology of the most abundant aboveground arthropods that inhabit maize crops in Central Spain. We have

75

76

examined the activity-density patterns of spiders, ground beetles and rove beetles collected in commercial Bt maize and non-Bt maize in Central Spain.

Materials and Methods This study was conducted in a large commercial maize field in the province of Madrid (Central Spain), during three consecutive years, from 2000 to 2002. Epigeal fauna of Bt maize fields (Bt+) (event 176, Compa CB; Syngenta Seeds SA) was compared with that of its near isogenic line without insecticide seed-treatment (Bt-) (NK Dracma; Syngenta Seeds SA) and with imidacloprid insecticide seed-treatment (Gaucho 35 FS, Bayer) (ImBt-). The planting dates were 13th April 2000, 19th April 2001 and 9th April 2002. The three different treatments were contrasted by a randomized block design with three replicates each. The total surface of the study area was ~ 5.0 ha. The fields were maintained according to the conventional farm practices, but without spraying insecticide.

Pitfall traps were used to collect aboveground arthropods. Each trap consisted of a plastic cup (12.5 cm diameter x 12 cm depth), with a funnel fitted in its upper rim, and both flush with the ground surface. Under the funnel a plastic container of 150 cc, where the insects fell down, was half filled with a 3:1 mixture of water and ethanol. Five traps per plot were placed, running a diagonal line along the plot, starting at least 6 m far from the field boundary to avoid edge effect. Traps were operative for three days every two weeks and the sampling period lasted from mid June to the end of September, giving a total of eight sampling dates per year. They were taken to the laboratory, and the arthropods were categorized by species.

Richness and diversity indices were calculated for spiders, ground beetles and rove beetles. The Shannon-Wiener diversity index (H’) was computed to detect changes in the community structure of the predominant groups among the three different treatments. It was calculated with the following formula (Magurran, 1989):

H’ = - ∑ pi ln pi being pi the proportion of individuals found in the ith species. Differences on species richness and diversity among the three kinds of plots and years were tested by two-way ANOVA analyses. If overall differences among them were detected, pairwise comparisons were made by Newman-Keuls test. Results and Discussion Arthropod taxa in the maize crops Of all arthropods collected in the 45 pitfall traps during the three years of the study, 92% were spiders, ground beetles and rove beetles. All of them are generalist predators commonly found in maize (Lang et al., 1999; Albajes et al., 2003; De la Poza et al., 2005), and the two former are probably the most abundant components of the aboveground fauna in farming systems (Melnychuck et al., 2003). Activity-density patterns Temporal activity-density patterns of these three groups fluctuated from June to September (Farinós et al., 2008). The dynamics of spiders and ground beetles greatly depended on the year, showing peaks in different sampling dates with no relationship with the treatment. Only in the case of rove beetles a higher number of individuals was collected in Bt- plots with

77

respect to the other two treatments in the first sampling date of 2000 and 2002. Pollen shedding was mostly in mid-July, but no drastic changes in activity densities were observed in any of the non-target groups in Bt plots after anthesis. The great temporal variation found in these groups’ dynamics within a year and through the years regardless the treatment indicated that different factors further than the use of transgenic plants could affect their populations. Indicators of community structure of spiders, ground beetles and rove beetles Of the 37 species of spiders collected in the three years 96% belonged to six species, and this percentage was comparable every year in the three treatments. No differences in density of these species among Bt+, Bt- and ImBt- plots were detected. Mean richness ranged between 11.0 and 17.0 species, though significant differences among treatments were not found. The diversity of species was higher in 2000 than in 2001 and 2002, but no significant differences among the treatments were found.

Thirty-two species of ground beetles were collected in the pitfall traps between 2000 and 2002, and only three species accounted for 93% of the total, but no significant differences were found among treatments for these species when analysed separately. Mean richness values were lower than in spiders, ranging between 6.3 and 12.7 species, although no significant differences among treatments in each year were detected. Similarly, the Shannon-Wiener diversity index calculated every year did not show differences among treatments.

The third prevalent group was rove beetles with 32 species collected in the pitfall traps, three of them accounting for 59% of the total. In general, the species composition of rove beetles was very variable in the three years. Significant differences in the mean richness of rove beetles were observed, due to the lower number of species in ImBt- plots in relation to Bt+ and Bt- plots. In contrast, diversity was not significantly different in any of the treatments.

The absence of significant differences in species diversity indices calculated for spiders, ground beetles and rove beetles in transgenic and non-transgenic plots is consistent with results obtained in other field studies in Italy (Lozzia & Rigamonti, 1998; Lozzia, 1999).

It seems that features related to the own crop system rather than the Bt toxin were the major responsible of the composition of the aboveground community of arthropods and for the variation recorded within and between years. Acknowledgements We thank Javier Lahoz for technical assistance, and Ismael Sánchez-Ramos and Mª Teresa Moñivas (CIB, CSIC) for their help in field sampling. We are indebted to Clara Ruiz for her assistance ordering and sorting out all the rove beetles. We are also grateful to Ildefonso Ruiz-Tapiador, Jesús Luna, Antonio Melic and Raimundo Outeruelo for helping us in determining ground beetles, wolf spiders, two species of spiders and rove beetles, respectively. This work was supported by grants from the Spanish Ministry of Environment and the European Commission (FP6-502981). Gema P. Farinós has a contract of the program I3P of the CSIC, funded by the European Social Fund. References Albajes, R., López, C. & Pons, X. 2003: Predatory fauna in cornfields and response to

imidacloprid seed treatment. Journal of Economic Entomology 96: 1805-1813.

78

De la Poza, M., Pons, X., Farinós, G.P., López, C., Ortego, F., Eizaguirre, M., Castañera, P. & Albajes, R. 2005: Impact of farm-scale Bt maize on abundance of predatory arthropods in Spain. Crop Protection 24: 677-684.

Farinós, G.P., De la Poza, M., Hernández-Crespo, P., Ortego, F. & Castañera, P. 2008: Diversity and seasonal phenology of the aboveground arthropods in conventional and transgenic maize crops in Central Spain. Biological Control 44: 362-37.

Harwood, J.D., Wallin, W.G. & Obrycki, J.J. (2005) Uptake of Bt endotoxins by nontarget herbivores and higher order arthropod predators: molecular evidence from a transgenic corn agroecosystem. Molecular Ecology 14: 2815-2823.

Hilbeck, A., Baumgartner, M., Fried, P.M. & Bigler, F. 1998: Effects of transgenic Bacillus thuringiensis corn-fed prey on mortality and development time of immature Chrysoperla carnea (Neuroptera: Chrysopidae). Environmental Entomology 27: 480-487.

Lang, A., Filser, J. & Hensche, J.R. 1999: Predation by ground beetles and wolf spiders on herbivorous insects in a maize crop. Agriculture, Ecosystems and Environment 72: 189-199.

Lozzia, G.C. 1999: Biodiversity and structure of ground beetle assemblages (Coleoptera: Carabidae) in Bt corn and its effects on non target insects. Bolletino di Zoologia Agraria e di Bachicoltura Ser II 31: 37-58.

Lozzia, G.C. & Rigamonti, I.E. 1998: Preliminary study on the effects of transgenic maize on non target species. IOBC Bulletin 21: 171-180.

Magurran, A.E. 1989: Diversidad ecológica y su medición. Vedrá (ed.), Barcelona. Melnychuk, N.A., Olfert, O., Youngs, B. & Gillott, C. 2003: Abundance and diversity of

Carabidae (Coleoptera) in different farming systems. Agriculture, Ecosystems and Environment 95: 69-72.

Pilcher, C.D., Rice, M.E. & Obrycki, J.J. 2005: Impact of transgenic Bacillus thuringiensis corn and crop phenology on five nontarget arthropods. Environmental Entomology 34: 1302-1316.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 79-84 Changes in biochemistry of cucumber carrying the thaumatin II gene: relevance to herbivores Małgorzata Kiełkiewicz1, Janina Gajc-Wolska2, Maria Szwacka3, Stefan Malepszy3 1Department of Applied Entomology, 2Department of Vegetable and Medicinal Plants, 3Department of Plant Genetics, Breeding and Biotechnology, Warsaw University of Life Sciences, 02-776 Warsaw, Nowoursynowska 159, Poland (E-mail: malgorzata_kielkiewicz@sggw.pl) Abstract: The study evaluates the consequence of the thaumatin II gene insertion for nutritional suitability of GM-cucumber plants to insect/mite herbivores. Biochemical analyses revealed that among transgenic cucumber lines the constitutive level of leaf and fruit biochemical components differs. Therefore, the nutritional value of leaves and fruits of transformed and non-transformed cucumber lines was determined by calculating the ratio of primary (glucose, fructose, soluble protein) to secondary (methanol soluble phenolics; lignin) and secondary to primary compounds. Changes in the mutual proportions between primary and secondary metabolites seem to be relevant to insect/mite herbivore development as well as to consumers other than arthropods. Key words: thaumatin II gene, transgenic cucumber, leaf nutritional value index, insect/mite pests Introduction Cucumber plants carrying the thaumatin II gene have been modified to accumulate thaumatin II sweet protein to improve the taste of fruits (Szwacka et al., 2000; Szwacka et al., 2002). The suitability of some GM-cucumber lines for field production has been demonstrated (Gajc-Wolska et al., 2005).

Potentially, pest herbivores can be affected by thaumatin II per se as a family of pathogenesis-related (PR) proteins (Van Loon, 1997) and indirectly by changes of specific ratios of generally occurring plant compounds essential for pest development.

The nutritional requirements of herbivores have been largely determined in many feeding trials (Bernays & Chapman, 1994; Smith, 2005; Schoonhoven et al., 2005). Dietary protein is usually a source of the adequate amino acids. Carbohydrates like monosaccharides (glucose, fructose) or oligosaccharides (sucrose) greatly enhanced the growth of phytophagous pests. Nutritional quality of plants as a food source for pests is also modified by secondary compound phytochemistry of host plant.

It was recently shown that as a result of thaumatin II gene insertion into cucumber genome the level of some metabolites was altered (Tagashira et al., 2005; Kielkiewicz et al., 2006b, 2006c). Little is still known about the consequence of the thaumatin II gene insertion on nutritional suitability of GM-cucumber plants to insect and mite herbivores by modification of the level of some primary and secondary compounds. Materials and Methods Four lines (T224 09, T225 03, T212 01, T210 06) of GM cucumber and non-GM inbred line of Cucumis sativus L. Borszczagowski (line B) were cultivated under field conditions. Among

79

80

the chosen lines the level of thaumatin II varied being relatively high in leaves and fruits of line T224 09 and T212 01 (Szwacka et al., 2002).

Samples of leaves (3rd from the top of the main stem) and fruits (6-8cm long) were collected from 8 week old plants of transformed and untransformed (=control) plants and lyophilized. Soluble proteins (Bradford, 1976), reducing sugars (glucose and fructose, Nelson, 1944), methanol soluble phenolics (Jonson & Schaal, 1957) and lignin (Johnson et al., 1961) concentrations were assessed in leaf and fruit samples. The nutritional value of leaves and fruits of GM and non-GM cucumber lines was determined by calculating the ratio of protein to sugars, protein+sugars to phenolics, protein+sugars to lignin and phenolics to lignin.

The density of commonly appearing cucumber pests such as aphids, thrips and spider mites was counted 4 times during the early season and expressed as a mean number of specimens per leaf. Results and Discussion It has been shown that the concentration of leaf soluble protein in transformed cucumber lines T224 09 and T212 01 was lower than in the non-transformed (control) and the two other lines (T210 06, T 225 03) (Table 1). However, there were no statistically significant differences at the 95% confidence level. Table 1. The concentration of chemical compounds in leaves and fruits of 8 week old GM and non-GM cucumber plants growing in the open field. Treatment means were compared (ANOVA) and separated by parametric Tukey’s HSD test or non-parametric Kruskal-Wallis test at P=0.05. Means ± SD followed by different letters within the column are significantly different.

Treatment

Soluble protein (n=4)

(mg x g-1 DW)

Reducing sugars (n=4)

(mg x g-1 DW)

Phenolics (n=4)

(mg x g-1 DW)

Lignin (n=3) (%)

Leaf: B

T210 06 T225 03 T212 01 T224 09

27.50±5.55 27.38±3.79 26.50±3.08 19.38±5.56 24.25±3.50

20.94±1.39 16.38±1.11 19.81±2.17 20.06±2.68 19.00±3.76

18.57±0.27 c 17.75±0.20 b 16.33±0.01 a 18.64±0.22 c 19.71±0.14 d

3.47±0.41 b 2.48±0.60 a 2.42±0.34 a 2.50±0.40 a

3.14±0.17 ab Test statistic H = 5.6762

P = 0.2247 H = 7.4259 P = 0.1150

H = 17.8415 P = 0.0013

F4, 10 = 4.10 P = 0.032

Fruit: B

T210 06 T225 03 T212 01 T224 09

17.29±0.93 b 16.01±1.75 b 16.93±1.18 b 11.52±0.71 a 9.55±1.74 a

231.45± 7.07 204.03±34.39 227.65±12.70 219.32±31.57 208.81± 6.65

4.17±0.05 d 3.36±0.03 b 2.42±0.10 a 3.67±0.19 c 6.19±0.22 e

3.33±0.08 a

4.12±0.99 ab 3.42±0.78 a 5.49±0.12 b

3.95±0.98 ab Test statistic F4, 15 = 27.95

P ≤ 0.0001 H = 6.8286 P = 0.1452

F4, 15 = 404.86 P ≤ 0.0001

F4, 10 = 4.36 P = 0.0269

81

As compared to the control, the concentration of soluble protein in fruits of transformed lines T224 09 and T212 01 was distinctly diminished by 45 and 34%, respectively (Table 1). Thus, GM-cucumber lines with relatively high levels of thaumatin II (T224 09 and T212 01), (Szwacka et al., 2002) contained a relatively low level of soluble protein both in the leaves and fruits. Thagashira et al. (2005) showed that thaumatin II gene insertion into genome of cucumber plants triggered metabolomic changes such as decrease in the level of some amino acids.

The concentration of reducing sugars (glucose + fructose) in leaves and fruits of lines T225 03, T212 01 and T224 09 was not significantly different from the concentration in the control (Table 1). However, in fruits of all examined GM-cucumber lines a decreased level of reducing sugars compared to the control was demonstrated (Table 1).

After transformation, the concentration of leaf phenolics decreased by 4 and 12% in line T210 06 and T225 03, respectively. In line T224 09 it increased by 6% and there was no difference in line T212 01 compared to the control (Table 1). An increase in the level of benzoic acid in leaves of GM-modified cucumber lines compared to the parental line (control) was found by Thagashira et al. (2005). In fruits of cucumber lines carrying the thaumatin II gene the level of phenolics declined by 19, 42 and 12% in T210 06, T225 03 and T212 01 lines, respectively and increased by 48% in line T224 09 compared to the control (Table 1).

The concentration of lignin in leaves of GM-cucumber lines T210 06, T225 03 and T212 01 significantly differed compared to the control with exception of line T212 09, where the lignin level was not significantly reduced (Table 1). A decreased level of leaf lignin suggests a lower lignification of leaf cell walls in transgenic cucumber lines. The lignin concentration in fruits of transformed lines T210 06, T225 01 and T224 09 was similar to the lignin level in the control plants (Table 1). Only in fruits produced by line T212 01 the lignin amount was significantly increased (by 65%), which suggests an elevated lignification of fruit cell walls. The nutritional evaluation of transgenic cucumber fruits showed that transgenicity caused a sweeter taste of fruits, higher levels of protein as well as lower contents of Na, K, Ca, Mg and fibre in dry matter (Kosieradzka et al., 2001; Gajc-Wolska et al., 2005).

As a result of changes in phytochemistry of cucumber lines expressing the thaumatin II gene, leaf and fruit tissues differ in the proportion in which primary (growth-sustaining) and secondary (defence-related) metabolites occur (Table 2).

In both leaves and fruits of line T212 01 and T224 09, the ratios of soluble protein to reducing sugars were lower than those in control line B and transgenic lines T210 06 and T225 03 (Table 2). As compared to the control, the ratios of soluble protein and reducing sugars to phenolics in leaves of transgenic lines were decreased with the exception of line T225 03 and increased in fruits with the exception of line T224 09 (Table 2). The ratio of protein and reducing sugars to phenolics in leaves of transgenic lines T210 06 and T225 03 were stable with exception of line T212 01 and T224 09 showing a lowered ratio. The ratio of protein and reducing sugars to phenolics in fruits of transformed cucumber lines T210 06, T225 03 and T212 01 was increased by 10, 69 and 5%, respectively. In contrast, this ratio in fruits of line T224 09 was reduced by 46% (Table 2).

As compared to the control, the ratio of phenolics to proteins and reducing sugars was elevated in leaves of transgenic lines T210 06, T212 01, T224 09 by 8, 24 and 18%, respectively (Table 2). Only in leaves of T225 03 the ratio was reduced by 8% compared to the control. In fruits of lines T210 06, T225 03 and T212 01, the ratio of phenolics to protein and reducing sugars were lowered by 9, 41, and 5% compared to the control.

82

Table 2. The values of the ratio of primary to secondary and secondary to primary metabolites concentrations in leaves and fruits of transformed and non-transformed cucumber plants (based on the data in Table 1).

Treatment P : RS (P+RS) : Ph (P+RS) : L Ph : (P+RS) Ph : L

Leaf: B

T210 06 T225 03 T212 01 T224 09

1.31 1.61 1.34 0.97 1.27

2.61 2.46 2.84 2.12 2.19

14.04 17.65 19.29 15.65 13.95

0.38 0.41 0.35 0.47 0.46

0.54 0.72 0.68 0.74 0.64

Fruit: B

T210 06 T225 03 T212 01 T224 09

7.47 x 10-2 7.85 x 10-2 7.44 x 10-2 5.25 x 10-2 4.57 x 10-2

59.65 65.49 101.10 62.89 35.28

7.47 5.33 7.40 4.20 5.60

16.76 x 10-3 15.27 x 10-3 9.89 x 10-3 15.89 x 10-3 28.34 x 10-3

0.13 0.08 0.07 0.07 0.17

P – Soluble protein; RS – reducing sugars (glucose+fructose); Ph – methanol soluble phenolics; L – lignin

In contrast, this ratio in fruits of line T224 09 was increased by 69% compared to the control (Table 2).

Increased ratio of phenolics to lignin in leaves of all transformed cucumbers compared to the control suggests an altered balance between soluble and non-soluble secondary compounds (Table 2). In fruits of GM-cucumber lines decreased ratio of phenolics to lignin was noticed with exception of line T224 09 (Table 2).

Our studies demonstrated that the populations of both insect species - onion thrips (Thrips tabaci Uzel) and cotton aphids (Aphis gossypii Glover) and two-spotted spider mites (Tetranychus urticae Koch) were affected by the GM-cucumber lines (Table 3). Significantly lower numbers of arthropod-pests were recorded on GM-cucumber lines than on plants of line B (control) (Table 3). It is well known that herbivores are able to select and feed based on a specific ratio of generally occurring plant compounds (Bernays & Chapman, 1994; Smith, 2005; Schoonhoven et al., 2005). Thus, the variation in leaf quality of transformed cucumber lines described here could have a significant impact on the performance and development of common cucumber arthropods. However, deterrence (antixenosis) and/ or toxicity (antibiosis) can provide explanations for decreased development rates of examined herbivores on GM-cucumber plants. Other results of a recent field trial (Kielkiewicz et al., 2006a, b, c) showed that transgenic cucumber lines expressing the thaumatin II gene affected the abundance of some piercing-sucking pests and had no negative effect on natural enemies.

83

Table 3. The density of cotton aphids (Aphis gossypi Glover), onion thrips (Thrips tabaci communis Uzel) and two-spotted spider mite (Tetranychus urticae) (avg. number of specimens per leaf sample). Values shown are means ± SD from 4 replicates (fields) with 8-10 leaf samples per treatment (line) and harvest. Significant differences (P=0.05 using non-parametric Kruskal-Wallis test) among treatments in a group are indicated by different letters.

Treatment

Cotton aphid

n=4

Onion thrips

n=4

Two-spotted spider mites

n=4 B 19.0±11.0 b 3.1±1.5 b 3.1±1.2 b

T210 06 1.3±0.7 a 1.6±0.9 ab 0.7±0.5 a T225 03 1.1±0.4 a 1.9±0.9 ab 1.2±0.2 a T212 01 1.8±1.1 a 1.8±0.5 ab 1.3±1.0 a T224 09 1.3±0.4 a 1.0±0.1 a 0.4±0.2 a

Test statistic H = 10.7850 P = 0.0291

H = 9.4535 P = 0.0507

H = 13.2263 P = 0.0102

In summary: 1. A change in the level of both leaf and fruit primary/secondary components in GM-

cucumber lines with a relatively high level of thaumatin II suggests alternations in biosynthetic pathways.

2. A change in specific ratios of generally occurring leaf cucumber compounds could affect the preference of naturally occurring arthropod pests and have an impact on their further development rates.

3. In fruits of transformed cucumber lines, changed ratios of secondary to primary and primary to secondary metabolites together with the thaumatin II presence might result in a different nutritional quality of fruits for consumers other than arthropods (snails, birds, humans).

In our future research, we aim to determine, whether unexpected metabolic changes in

plant organs like leaves and fruits of GM-cucumber lines will exist in subsequent generations.

Acknowledgements This work was supported by grant No. P06R 017 29 from the Ministry of Science and High Education, Poland. References Bernays, E.A. & Chapman, R.F. 1994: Host-Selection by Phytophagous Insects. Chapman and

Hall, London. 305 pp. Bradford, M.M. 1976: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Ann. Biochem. 72: 248-254. Gajc-Wolska, J., Szwacka, M. & Malepszy, S. 2005: The evaluation of fruit quality (Cucumis sativus L.) of transgenic lines with thaumatin gene. Folia Hort. 17/2: 23-28. Johnson, D.B., Moore, W.E. & Zank, L.C. 1961: The spectrophotometric determination of lignin in small wood samples. TAPPI 44: 793-798.

84

Johnson, G. & Schaal, L.A. 1957: Accumulation of phenolic substances and ascorbic acid in potato tuber tissue upon injury and their possible role in disease resistance. Am. Potato J. 34: 200-209. Kiełkiewicz, M., Gajc-Wolska, J., Szwacka, M. & Maleszy S. 2006a: Insect and mite pests

and their natural enemies on genetically modified cucumbers expressing the thaumatin II gene [Szkodniki i fauna pożyteczna na transgenicznych ogórkach z ekspresją genu taumatyny II]. Progress in Plant Protection/Postępy w Ochronie Roślin 46(2): 457-469.

Kiełkiewicz, M., Szwacka, M., Gajc-Wolska, J. & Maleszy, S. 2006b: Genetically modified cucumbers with thaumatin II gene expression and their acceptance by pests

[Transgeniczne ogórki z ekspresją genu taumatyny II ich akceptacja przez szkodniki]. Advances of Agricultural Science Problem Issues [Zeszyty Problemowe Postępów Nauk Rolniczych] 509: 395-404.

Kiełkiewicz, M., Gajc-Wolska, J., Szwacka, M. & Maleszy, S. 2006c: Impact of transgenic cucumbers expressing the thaumatin II gene on the occurrence of arthropod fauna.

IOBC/wprs Bull.: Proceeding of the meeting ‘Breeding for inducible resistance against pests and diseases’ at Heraklio, Crete, 27-29 April, 2006 (in press).

Kosieradzka, I., Sawosz, E., Pastuszewska, B., Szwacka, M., Malepszy, S., Bielecki, W. & Czuminska, K. 2001: The effect of feeding diets with genetically modified cucumbers on

the growth and health status of rats. J. Animal Sci. 10: 7-12. Nelson, N. 1944: A photometric adaptation of the Somogyi method for determination of glucose. J. Biol. Chem. 153: 375-380. Schoonhoven, L.M., van Loon, J.J.A. & Dicke, M. 2005: Insect-Plant Biology. Oxford

University Press Inc., New York, 421 pp. Smith, C.M. 2005: Plant resistance to Arthropods. Molecular and Conventional Approaches. Springer, Dordrecht, The Netherlands, 423 pp. Szwacka, M., Krzymowska, M., Kowalczyk, M.E. & Osuch, A. 2000: Transgenic cucumber

plantsexpressing the thaumatin gene. Progress in Biotechnology. Food Biotechnology (S. Bielecki, J. Tramper, J. Polak, eds) Elsevier Science B.V., pp. 43-48. Szwacka, M., Krzymowska, M., Osuch, A., Kowalczyk, M.E. & Malepszy, S. 2002: Variable properties of transgenic cucumber plants containing the thaumatin II gene from

Thaumatococcus daniellii. Acta Physiol. Plantarum 24: 173-185. Tagashira, N., Pląder, W., Filipecki, M., Yin, Z., Wiśniewska, A., Gaj, P., Szwacka, M.,

Fiehn, O., Hoshi, Y., Kondo, K., Malinowski, R. & Malepszy, S. 2005: The metabolic profiles of transgenic cucumber lines vary with different chromosomal locations of the transgene. Cell Mol. Biol. Lett. 10: 697-710.

Van Loon, L.C. 1997: Induced resistance in plants and the role of pathogenesis-related proteins. Eur. J. Plant Pathol. 103: 753-765.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 85-92 Belowground volatile emission of Bt maize after induction of plant defence Michael Meissle1,2, Ivan Hiltpold2, Ted C. J. Turlings2, Jörg Romeis1

1 Agroscope Reckenholz-Tänikon Research Station ART, Reckenholzstrasse 191, 8046 Zürich, Switzerland (E-mail: michael.meissle@art.admin.ch);2 University of Neuchâtel, Institute of Biology, Case postale 158, 2009 Neuchâtel, Switzerland Abstract: Roots of maize plants attacked by larvae of the Western corn rootworm (Diabrotica virgifera virgifera) commonly emit β-caryophyllene. This volatile sesquiterpene has been shown to attract entomopathogenic nematodes, potential biological control agents against this serious pest. Transgenic maize expressing coleopteran-specific insecticidal Cry3Bb1 protein from Bacillus thuringiensis (Bt) is another available control option. To improve efficacy of Bt maize and reduce the probability of resistance development, a combination of both methods could lead to a more sustainable system, as Bt maize kills or deters neonate larvae and nematodes would be capable of infecting older larvae surviving the Cry3Bb1 exposure. In the present study, we examined the emission of β-caryophyllene of two different Bt lines and their corresponding non-transformed near isolines after induction by a bacterial elicitor or by D. v. virgifera larvae. All maize lines were similarly capable of emitting β-caryophyllene after roots were incubated with coronatine, although variation was high. Emission after feeding by D. v. virgifera larvae was lower and less reliable compared with coronatine. The two Bt maize/control lines showed rather low β-caryophyllene levels compared to a highly attractive control line. Whether or not the emission levels in the varieties tested would be enough for successful nematode attraction and consequently effective biological control remains to be shown. Key words: Bt maize, plant defence, caryophyllene, coronatine, Diabrotica virgifera virgifera Introduction The western corn rootworm (WCR, Diabrotica virgifera virgifera, Coleoptera: Chrysomelidae) threatens European maize production since its introduction in the late 1980s. Environmentally friendly control methods are needed to control this pest and biological control is one potential option. Several species of entomopathogenic nematodes (EPN) which are capable of attacking and killing WCR larvae feeding on maize roots have been identified (Kuhlmann & van der Burgt, 1998; Toepfer et al., 2005). Such nematodes are available as commercial biological control products for the control of other insect pests and may be used effectively for WCR control when mass-reared and applied appropriately. Another promising option to control WCR is transgenic maize expressing insecticidal, coleopteran-specific, proteins derived from Bacillus thuringiensis (Bt) (Vaughn et al., 2005). Besides Bt maize expressing toxins against stem boring Lepidoptera, which is grown on large areas worldwide for the last decade, Bt maize against corn rootworms has entered commercial production in 2003 and is grown on steadily increasing acreages ever since, mainly in the USA. Integrated pest management (IPM) promotes the combination of environmentally friendly pest reduction methods instead of stand-alone techniques with the aim of reducing overuse of chemical. Growing Bt maize events, which do not provide full protection against WCR larvae (Al-Deeb & Wilde, 2005), combined with the a biological control agent, such as EPN, may be a more sustainable approach to control WCR, which might also delay resistance development

85

86

(Gassman et al., 2006). A prerequisite for successful biological control by EPN is that nematodes are able to locate and find their hosts. This requires a functioning plant-nematode communication.

When plants are attacked by herbivores, they have been found to emit a specific spectrum of highly volatile organic compounds. Natural enemies are attracted by tracking the source of those chemicals. This has been shown for a number of aboveground systems, mainly predatory mites and parasitic wasps (Dicke & Vet, 1999; Turlings & Wäckers, 2004). Recent research has revealed that volatile emission is also induced belowground after WCR attack in maize roots (Rasmann et al., 2005).The composition of the emitted volatiles however may be influenced by a number of factors, e.g. the genetic background, as shown for maize. While most European maize varieties initiate the emission of the sesquiterpene β-caryophyllene after herbivore attack, American maize lines appear to have lost this specific signal (Degen et al., 2004; Rasmann et al., 2005). Transgenic maize, transformed to express Cry1Ab (N4640Bt, transformation event Bt 11) against stem boring Lepidoptera emitted a reduced amount of volatiles when induced by caterpillar regurgitant when compared with non-transformed plants. Despite this fact, the attractiveness to parasitoid wasps seemed to be unaffected (Turlings et al., 2005), possibly because unknown key attractants, which are likely to be released in small amounts, remained unaffected (D’ Alessandro & Turlings, 2006). In contrast, Dean & De Moraes (2006) reported similar volatile emission patterns for Cry1Ab expressing transgenic maize (DKc6125, transformation event MON 810) when damage levels were controlled to be similar to the corresponding non-transgenic line. Changes in herbivore-induced volatile profiles of Bt maize were discussed as a consequence of altered feeding behaviour rather than of changes in biochemical plant defence pathways.

The belowground volatile spectrum emitted by WCR attacked maize roots seemed to be reduced to a few compounds, i.e. α-humulene, β-caryophyllene and caryophyllene oxide, of which β-caryophyllene was the compound most prominently emitted (Rasmann et al., 2005). In laboratory and field experiments, β-caryophyllene attracted Heterorhabditis megidis, a nematode species that is well capable of infecting WCR larvae (Rasmann et al., 2005). For those reasons, we focused on β-caryophyllene in the present work. For the transgenic Bt maize varieties available today, it is unclear whether plants emit belowground volatiles after attack and if such emission is comparable to the non-transformed counterpart lines. This would be an important prerequisite for successful combination of Bt maize with biological control by nematodes.

Cry3Bb1 expressing Bt maize has been developed to reduce feeding by D.v. virgifera larvae by killing or deterring them from feeding. As a direct consequence of reduced feeding, lower induction of plant defence volatiles and thus natural enemy attraction is expected. Schuler et al. (1999) have shown a similar relationship for aboveground parasitoid attraction in Bt oilseed rape. In order to exclude the factor of different feeding rates, several methods could be considered: (1) induce with strains or species resistant to the expressed insecticidal protein; (2) induce with larval spit (regurgitant) of the herbivore instead of the herbivore itself (Turlings et al., 2005); (3) induce artificially by using chemical inducing agents (elicitors). We decided to use elicitors for our experiments, as no appropriate resistant species or WCR spit collection methods were available. The elicitor chosen was coronatine, a phytotoxic chemical produced by the bacterium Pseudomonas syringue, which triggers defence responses when applied to plant tissue (Boland et al., 1995). In the present study, we compared the volatile spectrum of elicitor- and D. v. virgifera-induced Bt maize roots with that of near isogenic control maize roots using gas chromatography-mass spectrometry. Implications of our findings for biological control with entomopathogenic nematodes are discussed.

87

Material and Methods Plants Two different Bt/control maize lines (provided by Monsanto Co., St. Louis, USA) were used in the experiments: 1) DKc5143 expressing Cry3Bb1 protein (transformation event MON88017) against corn rootworms and the corresponding non-transformed near-isoline (control) DKc 5143 – maize varieties carrying this transformation event are not commercialized yet. 2) DKc 3420 expressing Cry1Ab protein (transformation event MON810) against stem boring Lepidoptera and the corresponding background variety DKc 3420 – This transformation event was bred into many varieties marketed worldwide including European countries like Spain and France. Both conventional maize varieties DKc5143 and DKc3420 are registered in the European Union and regarded as “European” by the breeders (Monsanto, personal communication). Thus they were expected to emit β-caryophyllene. In preliminary experiments, we ensured that β-caryophyllene is produced in at least some of the lines we wanted to test. The variety “Graf”, which is known to release considerable amounts of β-caryophyllene after induction by D. v. virgifera (Rasmann et al., 2005) was used as a reference. Plants were grown in 5cm plastic pots filled with seedling-soil in a phytotron climate chamber (23°C, 65% humidity, 16:8 light:dark conditions) and used 10-15 days after sowing. Coronatine induction in agar filled plates Purified coronatine from Pseudomonas syringae was purchased from Sigma (Buchs, Switzerland, LOT 07444139, Cat. no. C-8115 62251-96-1).

A fast and easy method of inducing volatile emission in roots is to apply diluted elicitor solution directly to the root system of maize seedlings. To ensure a constantly moist environment around the roots during induction, the lid and the bottom of specially prepared Petri-dishes (9cm) were filled with 1% water-agar. To half of the dishes, a piece of filter paper (5.5cm diameter) was added on which 20 μl coronatine solution (2.5nmol/μl) dissolved in 1ml deionized water was applied. The other half of the dishes was provided with filter paper and deionized water only. The root system of individual maize seedlings was washed carefully and the tips of the roots were cut to ensure better coronatine uptake. After that, each seedling’s roots were sandwiched in between a lid and bottom of the dish so that they were in close contact with the filter paper. A hole cut into the rim of the dishes allowed the stems and foliar parts of the maize plants to stick out of the sandwich. Tinfoil was then wrapped around the dishes to protect the roots from light and plants were left in the laboratory for 1 day at ambient conditions. For each of the 2 Bt/control maize pairs, 7-15 plants were induced with coronatine and the same number of plants remained not induced. Coronatine and D. v. virgifera induction in sand In order to compare coronatine induced roots with D. v. virgifera induced ones, we changed to a system using maize seedlings planted in moist sand. After the roots of maize seedlings were washed, 200ml glass pots (as described in Rasmann et al., 2005) were filled with sterilized white sand (Migros, Switzerland). For coronatine induction, each seedling was planted in a pot and 50nmol coronatine dissolved in 15ml deionized water were added directly with a glass pipette into the sand near the root system. In the control treatment, only water was applied. Alternatively, a cotton dental wick was soaked with 3ml deionized water containing 100nmol coronatine. The washed roots of each seedling were wrapped around the cotton and buried into the sand in the pots. Plants were induced under 14:10 light:dark and ambient temperature and humidity conditions for 2 days. For D. v. virgifera induction, after each

88

seedling was planted in a pot, 4-5 small holes were made into the surface of the sand near the plants. One larva in the second larval stage (L2) was added to each hole and the holes were covered gently with sand. Larvae were allowed to establish and feed for 5 days on the roots. For the latter experiments, we focussed on transgenic and non-transgenic DKc3420 maize plants expressing Cry1Ab protein, as we wanted to examine volatile emission under similar induction stimuli. Cry3Bb1 protein would likely have resulted in reduced root feeding and thus in a different induction of plants. Each of the described induction methods was replicated 3-4 times per maize line.

To compare the β-caryophyllene production levels of the two examined Bt/control maize pairs with previous results (Rasmann et al., 2005), we induced Graf (known to emit high β-caryophyllene levels when fed upon by D. v. virgifera larvae) and one of our control varieties (DKc5143) with D. v. virgifera larvae as described above, except that 10 larvae were used per plant and larvae were allowed to feed for 4 days only. For each treatment 8-9 plants were induced. Volatile collection and analysis After the incubation period in each experiment, roots were cleaned and ground in liquid nitrogen using mortar and pestle. After transferring the macerated tissue to airtight glass vials with a septum in the lid, the fresh weight of the sample material was recorded and the vials were stored at -75°C until further analysis. After adjusting to room temperature, vials were placed in an autosampling system. A 100mm polydimethylsiloxane solid phase micro extraction fibre (SPME, Supelco, Buchs, Switzerland) attached to the system was inserted through the septum of the vials and exposed for 20 min at 40°C. Thereafter, the fibre was inserted into the injector port (splitless inlet) of an Agilent 6890 Series GC system G1530A coupled to a quadrupole-typemass-selective detector (Agilent 5973; transfer line and source 230°C, ionization potential 70eV), where the compounds adsorbed on the fibre were desorbed at 230°C. Chromatography was conducted on an apolar column (HP-1MS, 30m length, 0.25mm internal diameter, 0.25μm film thickness). Helium at a constant pressure (128kPa, 0.9ml/min flow, 35cm/s average velocity) was used as carrier gas. After fibre insertion, the column temperature was maintained at 60°C for 1min, then increased by 20°C per minute until 250°C were reached and finally maintained at 250°C for another 10 min. Data recorded by the mass spectrometer were processed to chromatograms using Enhanced Chem Station G1701DA software (Version D.00.00.38, Agilent Technologies, 2001). Peaks were identified using an institute intern plant volatile library and the Mass Spectral Library Version NIST02 (National Institute of Standards and Technology, Gaithersburg, USA). The area under each peak was calculated and divided by the fresh weight of the root samples. Differences between varieties or induction methods were statistically analyzed using nonparametric Kruskal-Wallis tests or Mann-Whitney U-tests. Results and Discussion When induced with coronatine in agar-filled Petri-dishes over night, the elicitor was in closed contact with the bare roots of the maize seedlings. As a consequence, all plants in the coronatine treatment emitted β-caryophyllene (Figure 1). However, the level measured per g fresh weight of ground roots varied considerably from very low expression levels to relatively high levels. When comparing the 4 maize varieties tested, no differences were evident (Kruskal-Wallis test, p>0.05; Figure 1). Non-induced plants only occasionally emitted minor amounts of β-caryophyllene. The difference between induced and not induced plants (pooled for all varieties) was significant (Mann-Whitney U-test, p<0.05)

89

Figure 1. Emission of β-caryophyllene by maize seedling roots from control and transgenic plants (MON88017, MON 810) induced with coronatine or not induced in agar- filled Petri-dishes (N = 7-15 per treatment).

Figure 2. Emission of β-caryophyllene by maize seedling roots from control and Cry1Ab expressing transgenic plants DKc3420 (MON810) induced with coronatine (either 50nmol directly applied to the sand, N=3, or 100nmol applied in cotton wick, N=3), Diabrotica v. virgifera larvae (4-5 L2 larvae per pot, N=4) or not induced (N=6) in sand-filled glass pots.

A similar pattern was observed when plants were induced in sand-filled glass pots (Figure 2). There was no difference between coronatine induced Bt plants (DKc3420 MON810) and control plants (Mann-Whitney U-test, p>0.05). In this assay, coronatine was applied in a larger volume of water than in the Petri-dish experiment (3ml in the cotton wick or 15ml directly pipetted into sand) and the volume of the sand filled glass pots was considerably larger than that of the Petri-dish arenas. On the other hand, plants were incubated for 2 days instead of 1 day. Consequently, it is possible that the amount of coronatine reaching the roots was different than in the Petri-dish arenas. However, measured β-caryophyllene levels were similar in both assays. When plants were induced by 4 D. v.

90

virgifera larvae (L2) as previously done by Rasmann et al (2005), induction was significantly lower than with coronatine (Bt and control plants pooled, Mann-Whitney U-test, p<0.05). In fact, β-caryophyllene was only measurable in 3 out of 8 plants (Figure 2).

As was found by Rasmann et al. (2005), roots from the variety Graf emitted large amounts of β-caryophyllene after feeding by D. v. virgifera larvae. These amounts were significantly higher than what was emitted by the roots of DKc5143 seedlings when fed upon by D. v. virgifera larvae (Mann-Whitney U-test, p<0.05; Figure 3).

Figure 3. Emission of β-caryophyllene from maize seedlings from control plants DKc5143 (N=8) and Graf (N=9) induced with D. v. virgifera larvae and not induced control plants (N=8) in sand filled glass pots.

Figure 4. Typical chromatograms obtained for D. v. virgifera induced maize roots of the variety Graf and DKc5143 and for not induced DKc5143. Only a part of the 20min measurement time of the GC run is shown.

To illustrate the different levels of β-caryophyllene emission after induction by D. v. virgifera larvae, the chromatograms of a typical, induced Graf and DKc5143 plant and an uninduced plant are shown in Figure 4. In contrast to the Graf plant, which showed a pronounced β-caryophyllene peak at retention time 7.88min, the induced DKc5143 plant

91

emitted only little β-caryophyllene. No such peak was measurable for the uninduced plant (Figure 4). There is more evidence for low β-caryophyllene production of our Bt and control plants from the coronatine induction experiment in sand in which the elicitor (100nmol) was applied on a cotton wick. In addition to the plants presented in Figure 2, we induced Delprim plants (also shown to have high volatile emission levels, Rasmann et al., 2005). Very high β-caryophyllene emission levels were observed with a response peak area of ca. 90x106 units per g FW, which was more than 4 times higher than the highest value measure for DKc3420 MON810. Conclusions and Implications for Biological Control

1. The two maize lines DKc3420, DKc5143 and their two transgenic counterparts with

either the transformation event MON810 or MON88017, respectively, are all capable of emitting β-caryophyllene after induction of plant defence with the bacterial elicitor coronatine. There are no apparent differences between the Bt and non-transformed lines and between the varieties DKc3420 and DKc5143.

2. Caryophyllene emission triggered after root-feeding by D. v. virgifera larvae was lower and less reliable in comparison with chemical elicitor-induction. DKc5143 and DKc3420 showed much lower β-caryophyllene levels when subject to the same induction stimuli than the varieties Graf and Delprim, which have previously been shown to attract the entomopathogenic nematode Heterorhabditis megidis.

3. For biological control with entomopathogenic nematodes, a high emission of key attractants for the enemies of insect pests is desirable. Whether or not the emission levels of β-caryophyllene (and probably other volatiles) after D. v. virgifera induction in the varieties tested would be enough for successful nematode attraction and consequently effective biological control remains to be shown. Breeding higher volatile responses into commercial maize varieties (including lepidopteran resistant Bt maize) could improve the possibility of using entomopathogenic nematodes as effective biological control agents against root feeding Diabrotica spp. (Rasmann et al., 2005).

4. Considering that current Bt maize targeting D. v. virgifera available today do not kill the larvae completely, a combination with biological control could be a more sustainable approach in terms of efficacy and resistance management. Bt maize varieties with a high emission of nematode attracting volatiles could lead to improved host finding of the nematodes and may help to kill larvae that survive feeding on Bt maize. Acknowledgements We thank Monsanto Co. (St. Louis, USA) for providing transgenic and control maize seeds, Claudia Zwahlen for practical help and valuable comments on the manuscript and the National Center of Competence in Research (NCCR) Plant Survival research program of the Swiss National Science Foundation for kindly funding the project. References Al-Deeb, M.A. & Wilde, G.E. 2005: Effect of Bt corn expressing the Cry3Bb1 toxin on

western corn rootworm (Coleoptera: Chrysomelidae) biology. J. Kansas Entomol. Soc. 78: 142-152.

Boland, W., Hopke, J., Donath, J., Nüske, J. & Bublitz, F. 1995: Jasmonsäure- und Coronatin-induzierte Duftproduktion in Pflanzen. Angew. Chem. 107: 1715-1717.

92

D’Alessandro, M. & Turlings, T.C.J. 2005: In Situ modification of herbivore-induced plant odours: A novel approach to study the attractiveness of volatile organic compounds to parasitic wasps. Chem. Senses 30: 739-753.

Dean, J.M. & De Moraes, C.M. 2006: Effects of genetic modification on herbivore-induced volatiles from maize. J. Chem. Ecol. 32: 713-724.

Degen, T., Dillmann, C., Marion-Poll, F. & Turlings, T.C.J. 2004: High genetic variability of herbivore-induced volatile emission within a broad range of maize inbred lines. Plant Physiol. 135: 1928-1938.

Dicke, M. & Vet, L.E.M. 1999: Plant-carnivore interactions: evolutionary and ecological consequences for plant, herbivore and carnivore. In: Herbivores: between plants and predators, eds. Olff, Brown & Drent, Blackwell, Oxford, pp. 483-520.

Gassmann, A.J., Stock, S.P., Carriere, Y. & Tabashnik, B.E. 2006: Effect of entomopathogenic nematodes on the fitness cost of resistance to Bt toxin Cry1Ac in pink bollworm (Lepidoptera : Gelechiidae). J. Econ. Entomol. 99: 920-926.

Kuhlmann, U. & Van der Burgt, W.A.C.M. 1998: Possibilities for biological control of the western corn rootworm, Diabrotica virgifera virgifera LeConte, in Central Europe. Biocontrol News and Information 19: 59N-68N.

Rasmann, S., Kollner, T.G., Degenhardt, J., Hiltpold, I., Toepfer, S., Kuhlmann, U., Gershenzon, J. & Turlings, T.C.J. 2005: Recruitment of entomopathogenic nematodes by insect-damaged maize roots. Nature 434: 732-737.

Schuler, T.H., Potting, R.P.J., Denholm, I. & Poppy, G.M. 1999: Parasitoid behaviour and Bt plants. Nature 400: 825-826.

Toepfer, S., Gueldenzoph, C., Ehlers, R.U. & Kuhlmann, U. 2005: Screening of entomopathogenic nematodes for virulence against the invasive western corn rootworm, Diabrotica virgifera virgifera (Coleoptera: Chrysomelidae) in Europe. Bull. Ent. Res. 95: 473-482.

Turlings, T.C.J., Jeanbourquin, P.M., Held, M. & Degen, T. 2005: Evaluating the induced-odour emission of a Bt maize and its attractiveness to parasitic wasps. Trans. Res. 14: 807-816.

Turlings, T.C.J. & Wäckers, F. 2004: Recruitment of predators and parasitoids by herbivore-injured plants. In: Advances in insect chemical ecology, eds. Cardé & Millar, Cambridge University Press, Cambridge, pp. 21-75.

Vaughn, T., Cavato, T., Brar, G., Coombe, T., DeGooyer, T., Ford, S., et al. 2005: A method of controlling corn rootworm feeding using a Bacillus thuringiensis protein expressed in transgenic maize. Crop Sci. 45: 931-938.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 93-99

Assessment of possible non-target impacts of the novel Bt-maize event MON88017 resistant against the Western Corn Rootworm Diabrotica virgifera virgifera (LeConte) Stefan Rauschen, Ingolf Schuphan, Sabine Eber RWTH Aachen University, Institute of Environmental Research, Worringerweg 1, D-52074 Aachen, Germany (E-mail: Rauschen@Bio5.RWTH-Aachen.de) Abstract: The Western Corn Rootworm (WCR) Diabrotica virgifera virgifera (LeConte) is regarded as a major threat to European maize cultivation since its introduction to the Balkans in the early 1990s. So far it has spread very rapidly, reaching economically significant levels in Serbia, Croatia and Hungary.

The novel Bt-maize event MON88017, which expresses the coleopteran-specific protein Cry3Bb1, is resistant to WCR. We investigated possible non-target effects of this maize variety under field conditions. The occurrence and abundance of non-target arthropods, e.g. herbivores, pollen feeders, and generalist predators, was assessed in a 4 hectare field experiment with MON88017, its near isogenic line DKC5143 and the two conventional hybrids DK315 and Benicia. The four maize lines were planted in a randomized plot design with eight replicates each. Several sampling methods were tested for monitoring the abundance of non-target species.

Two arthropod species were selected for detailed studies, based on their prevalence and density in the field, and their exposure to Cry3Bb1 as documented by ELISA tests: the cicadellid leafhopper Zyginidia scutellaris (Herrich-Schäffer) and the mirid bug Trigonotylus caelestialium (Kirkaldy). Results from two cultivation periods did not show significant differences between the densities of both species in plots planted with Bt-maize and the near isogenic line. Significant differences exist, however, between these two hybrid lines and the two conventional hybrids. Analyses of the abundance data of T. caelestialium in terms of data gathered on the soil characteristics of individual plots showed no obvious influence of particular soil parameters. For Z. scutellaris the recorded densities varied strongly for different sampling methods and sampling dates. Key words: Diabrotica virgifera virgifera, non-target impact, risk assessment Introduction The Western Corn Rootworm (WCR) Diabrotica virgifera virgifera (LeConte) is regarded as a major threat to European maize cultivation. This chrysomelid was introduced to the Balkans in the early 1990´s and has since been continuously spreading, reaching economically significant levels in Serbia, Croatia and Hungary (Hummel, 2003). It is assumed that in a few years D. v. virgifera will have become established in most central European countries.

One strategy to control the population densities and economic impact of WCR is the use of genetically modified Bt-maize. Resistance to WCR is conferred by expression of the coleopteran specific Cry3Bb1 protein derived from Bacillus thuringiensis. For crop plants of this kind to be authorised, current legislation of the European Union stipulates a pre-market risk assessment and a post market monitoring (EC, 2001, EFSA, 2006).

We investigated possible non-target effects of the novel WCR-resistant Bt-maize transformation event MON88017 under field conditions. The occurrence and abundance of a functionally and systematically diverse range of non-target arthropods was assessed in this

93

94

transgenic maize hybrid in comparison to its near isogenic line and two conventional hybrids. Special emphasis was put on herbivorous arthropods which were prevalent and abundant in the field and proved to be exposed to the transgenic protein. Our objectives were to identify species which could be used as indicators for unintended non-target effects of transgenic maize, and to evaluate suitable methods to assess effects on these arthropods in the field. Material and Methods Experimental design Within an area of approximately 4 ha four different maize varieties were planted in a systematic plot design with 8 replicates each. Bt-maize MON88017 and its near isogenic line DKC5143 (both Monsanto Co.), and two conventional hybrids, Benicia (Pioneer HiBred, Johnston, Iowa, USA) and DK315 (Monsanto Co.), were used for the experiment. The plots measured 40.5 m by 31.5 m, yielding an area of 0.13 ha. The experimental field was surrounded by a 4.5 m clearance strip followed by a perimeter of conventional hybrid maize (Gavott, KWS Saat, AG, Einbeck, Germany) with at least 10 m width. In both study years, the locations of the plots with their respective maize varieties were identical. This ensured comparable abiotic parameters (i.e. soil structure, moisture content, pH-value etc.) over the whole experimental time for each replicate. Soil parameters Individual plots were characterised by a number of physico-chemical soil parameters (pH value; cation exchange capacity; per cent fractions of clay, sand and silt; carbon and nitrogen content; carbon/nitrogen ratio). Laboratory work was performed at the Institute of Applied Biotechnology in the Tropics, IBT e.V., Göttingen, Germany. Field sampling Based on our experience in a previous Bt-maize project and published literature, we chose two groups of arthropods as the main focus of research: the leaf- and planthoppers (Auchenorrhyncha) and herbivorous plant bugs (Heteroptera: Miridae). Sampling methods were chosen accordingly.

Densities of both focal arthropod groups were mainly recorded by sweep netting using a 40cm diameter net (mesh width 1.5 mm). Samples were taken in each plot in July and August 2005 (calendar weeks CW 28, 34), and in July, August, and September 2006 (CW 29, 33, 36). Four linear transects were covered in each plot, which divided the plots into five parts of similar size. At the end of each transect, captured insects were removed and stored in 70% ethanol.

For sampling Auchenorrhyncha, we also used custom made transparent sticky traps as described by Rauschen et al. (2004). One sticky trap was set up near the centre of each plot for one week after which the traps were covered with cling film and frozen. In the laboratory, Auchenorrhyncha were carefully removed from the sticky surface and stored in ethanol. Species identification was done under a binocular microscope. Exposure assessment Z. scutellaris has been shown to be exposed to novel proteins in transgenic maize (Dutton et al., 2004; Obrist et al., 2006) and its exposure has been confirmed for Cry3Bb1 from MON88017 with around 1 μg/g dry weight (M. Meissle, personal communication). Bulk samples and single specimens of T. caelestialium were analysed for their Cry3Bb1 content

95

with commercially available enyzme-linked immunosorbent assay (ELISA) test kits (Agdia Inc., Elkhart, IN, USA; purchased via Linaris GmbH, Wertheim-Bettingen, Germany).

Results and Discussion Two herbivores from the focal groups were the most abundant species in the experimental field: the maize leafhopper Zyginidia scutellaris Herrich-Schäffer (Cicadomorpha: Cicadellidae) and the rice leaf bug Trigonotylus caelestialium Kirkaldy (Heteroptera: Miridae). We concentrated our further analysis on those two species. Abundance of Z. scutellaris The maize leafhopper Z. scutellaris was present in high densities throughout the growing season of maize. Overall, the population densities in 2006 were markedly higher than in 2005. The two sampling methods reflected the population densities differently: While the sweep nets indicated an increased population in the later growing season, the sticky traps caught most insects early in the season (Figures 1 & 2). This was true for both study years. The difference between the number of individuals caught with the two methods increased during the season. This discrepancy may be attributed to a change in behaviour of Z. scutellaris and/or a difference in the ratios of males to females caught with the two methods (unpublished data). There were no significant differences between the leafhopper densities in Bt-maize plots versus those planted with the near isogenic line. Between the conventional lines, however, there were differences at certain sampling dates (not shown).

Z. scutellaris caught with sweep nets (2006)

0

100

200

300

400

500

600

700

July August September

tota

l of i

ndiv

idua

ls

BT MON88017ISOGENICDK315Benicia

Figure 1. Total number of Z. scutellaris adults caught with sweep nets during the 2006 season.

96

Z. scutellaris caught with sticky traps (2006)

0

50

100

150

200

250

300

July August September

tota

l of i

ndiv

idua

ls

BT MON88017ISOGENICDK315Benicia

Figure 2. Total number of Z. scutellaris adults caught with sticky traps during the 2006 season. Abundance of T. caelestialium The rice leaf bug was also present in high densities throughout the season (Table 1). In August, the emergence of a second generation could be observed in the maize field. Statistical analyses with general linear models (GLM) showed that there was an influence of the maize hybrids and of varying combinations of soil parameters on the densities of T. caelestialium (Table 2). No significant differences between the densities in MON88017 and the near isogenic line were detected. For different life stages of the rice leaf bug there were different density patterns in the four maize hybrids. While adults showed no consistent pattern early in the season, there were more nymphs in Benicia than in DK315, whereby MON88017 and its near isogenic line showed very similar, intermediate values (Figure 3). This pattern could be observed in both study years. The statistical significance of these results is still under investigation. Exposure of T. caelestialium Results show that this species is exposed to the Bt protein during all developmental stages: Adult females contained 17 ± 20 ng Cry3Bb1 per individual (N=13), while for adult males a “payload” of 3 ± 1 ng Cry3Bb1 (N=8) was measured with ELISA. In nymphs of both sexes comparable or even higher amounts were detected. These values correspond to concentrations of between 5 and 10 μg Cry3Bb1/g fresh weight, which equals 1/6 to 1/3 of the concentration in fresh leaf material of the Bt hybrid. In relation, this is more than has been previously reported for this species on MON810 (Obrist et al., 2006).

97

Table 1. Mean number (± standard deviation) of T. caelestialium adults and nymphs in different maize hybrids at different sampling dates (calendar week given in parenthesis); -: no sampling was carried out). July (CW 28 & 29) August (CW 33 & 34) September (CW 36)

year hybrid adults nymphs adults nymphs adults nymphs

2005 DK315 17.75 ± 8.41 0 1.38 ± 1.06 8.75 ± 3.99 - -

Benicia 13.63 ± 6.46 0 3.63 ± 2.33 29.75 ± 8.61 - -

MON88017 16.63 ± 8.31 0 3.25 ± 3.41 19.00 ± 6.76 - -

DKC5143 20.00 ± 8.86 0 2.63 ± 2.20 15.63 ± 2.83 - -

2006 DK315 20.88 ± 9.83 0.13 ± 0.35 1.50 ± 1.93 3.38 ± 4.34 1.00 ± 0.93 3.75 ± 1.28

Benicia 17.88 ± 10.38 0.63 ± 0.92 7.38 ± 5.24 18.38 ± 10.77 3.13 ± 1.55 8.88 ± 3.27

MON88017 31.13 ± 19.85 0.25 ± 0.46 7.38 ± 6.00 10.75 ± 6.56 1.13 ± 0.99 4.75 ± 2.60

DKC5143 37.25 ± 24.63 0.38 ± 0.52 5.25 ± 2.19 11.38 ± 8.47 1.33 ± 1.06 3.75 ± 3.11

Table 2. Amount of deviance (in percent) in the density data of T. caelestialium adults and nymphs explained by the different covariables used in general linear models (GLM). % deviance explained % deviance explained by hybrid by combinations of soil parameters

2005 July adults 26.80 39.20 Aug adults 20.02 25.76 Aug nymphs 36.40 - Aug nymphs 69.40 0.00

2006 July adults 9.85 44.10 Aug adults 47.60 9.28 Aug nymphs 44.94 24.35 Sept adults 38.40 - Sept nymphs 37.59 -

98

Figure 3. Mean densities over 8 plots of T. caelestialium adults (left) and nymphs (right) in the different maize hybrids at the indicated sampling dates Potential of Z. scutellaris and T. caelestialium as indicator organisms The maize leafhopper has been frequently reported from maize (Rauschen et al., 2004; Pons et al., 2005; Obrist et al., 2006) and is regarded as being bound to maize (Nickel, 2003). It is typical for maize fields throughout Europe. The rice leaf bug has also recently been documented in maize (Obrist et al., 2006). It has a large host range and a worldwide distribution (Wheeler et al., 1985). Both species occur in high densities throughout the growing season and are exposed to considerable amounts of transgenic protein.

Based on these characteristics, these two species can be regarded as candidate indicator organisms for potential unintended effects.

They may also play major roles in food-webs and thus as mediators of transgenic proteins to higher trophic levels: An analysis of the web contents of the generalist predator Theridion impressum (Araneae) showed that Auchenorrhyncha and Heteroptera are major parts of this spider´s diet (Árpás et al., 2005).

Acknowledgements We thank Dr. Frank Gessler, Dr. Sybille Pagel-Wieder and Heinz Hunfeld for the soil characterisation, the Bavarian State Research Center for Agriculture (LFL Bayern) for their cooperation and the Federal Ministry of Education and Research BMBF for financial support of the project (grant 0313279). References Árpás, K., Tóth, F. & Kiss, J. 2005: Foliage-dwelling arthropods in Bt-transgenic and isogenic

maize: a comparison through spider web analysis. Acta Phytopathol. Entomol. Hung. 40: 347-353.

99

Dutton, A., Obrist, L., D´Alessandro M., Diener, L., Müller, M., Romeis, J. & Bigler, F. 2004: Tracking Bt-toxin in transgenic maize to assess the risks on non-target arthropods. IOBC/wprs Bull. 27(3): 57-63.

EC (2001) Directive 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC. L 106/1 of 4/17/2001.

EFSA (2006) Guidance Document of the Scientific Panel on Genetically Modified Organisms for the risk assessment of genetically modified plants and derived food and feed. The EFSA Journal 99:1-100.

Hummel, H.E 2003: Introduction of Diabrotica virgifera virgifera into the Old World and its consequences: a recently acquired invasive alien pest species on Zea mays from North America. Commun. Agric. Appl. Biol. Sci. 68: 45-57.

Nickel, H. 2003: The Leafhoppers and Planthoppers of Germany (Hemiptera, Auchenorrhyncha), Patterns and strategies in a highly diverse group of phytophagous insects. Pensoft Series Faunistica, Sofia, Moscow.

Obrist, L.B., Dutton, A., Albajes, R. & Bigler, F. 2006: Exposure of arthropod predators to Cry1Ab toxin in Bt maize fields. Ecol. Entomol. 31: 1-12.

Pons, X., Lumbierres, B., López, C. & Albajes, R. 2005: Abundance of non-target pests in transgenic Bt-maize: A farm scale study. Eur. J. Entomol. 102: 73-79.

Rauschen, S., Eckert, J., Schuphan, I. & Gathmann, A. 2004: Impact of growing Bt-maize on cicadas: diversity, abundance and methods. IOBC/wprs Bull. 27(3): 137-142.

Wheeler, A.G. & Henry, T.J. 1985: Trigonotylus caelestialium (Heteroptera: Miridae), a pest of small grains: seasonal history, host plants, damage, and descriptions of adult and nymphal stages. Proc. Entomol. Soc. Wash. 87: 699-713.

100

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 101-104 Application of environmental risk assessments of pest resistant crops in different environments Jeremy B. Sweet Environmental Consultant, 6 The Green, Willingham, Cambridge CB24 5JA, UK (E-mail: jeremysweet303@aol.com) Abstract: Studying effects of pest and disease resistant crops on sensitive non-target organisms (NTOs) is a complex process and involves studies of life history traits of NTOs, such as reproduction, feeding biology and distribution, toxicity to different developmental stages and related features such as the regional distribution of GM pest resistant (PR) non-GMPR crops, other alternative host plants and their distribution in the landscape, in order to determine the exposure of the NTO. This paper proposes risk assessment methods that consider different receiving environments when trying to assess impacts on sensitive NTOs. Key words: Environmental risk assessment, non-target organisms, pest resistance, risk management Background Environmental Risk Assessments (ERA) of genetically modified (GM) pest and disease resistant (PR) crops naturally focus on the specificity of the PR genes, lethal and sub-lethal effects on non-targets and food chain effects. Studies conducted in the laboratory (e.g. tier 1 and 2 tests, Romeis et al., 2006) will identify potential risks to particular groups of species and hence to populations. Glasshouse and field experiments and monitoring of crops will identify probable risks to populations from exposure at field levels. However concerns remain about the effects of GMPR crops at the landscape scale and the effects that large scale continuous or rotational GMPR cropping may have on species identified as sensitive to PR gene products (Ferry et al., 2006). Some of these sensitive species may be rare with populations already threatened by agricultural practices and there are concerns that GMPR crops may add to the likelihood of localised or regional extinction. This paper reviews some of the recent work in this area and proposes risk assessment approaches for these situations. Risk Assessment for Different Regions When assessing risks to vulnerable species in a defined region the following steps in the ERA can be considered:

1. Study the specificity of the toxicity or bio-activity of the gene product (e.g. the tier 1 and 2 results) in relation to the species of concern in the region and consider which species are potentially at risk in a region.

2. The risk assessment considers the maximum exposure levels of the identified non-target organism to the GM plant and/or plant parts containing the biocidal product. These exposure levels should be compared with doses identified in laboratory tests as having lethal or sub-lethal effects in the species of interest.

101

102

3. The assessment considers the life cycle of these species and the development of vulnerable life stages (e.g. phytophagous larvae) on hosts (including the GMPR plants) in relation to the growth and development stages of the GMPR plant.

4. The distribution, spatial and temporal location of the host plant populations in the landscape of the region is determined.

5. The spatial and temporal distribution of the vulnerable life stages of the NTO on these host plants is determined in this landscape.

6. The spatial and temporal distribution of the GMPR plants/crops in the region is determined or predicted from current cropping practices.

7. From the above assessments, estimates can be made of the proportion of the vulnerable NTO population exposed to active levels of the biocide each season and in successive seasons.

Assimilating relevant data to conduct these studies may be problematic and thus modelling approaches can be considered. Certain assumptions may have to be made to develop parameters and weightings of parameters used in constructing the models. These need to be clearly stated so that the risk assessments can be fully evaluated by others. Experiences gained in other studies will help to inform the development of such models (Groot & Dicke, 2002).

Castellazzi et al. (2007) have developed spatial and temporal models for describing agricultural landscapes and their cropping patterns. For developing models of insect resistant (IR) GM crops, use can be made of numerous studies of IR crops on non-target arthropods (NTA) examining large numbers of species in different cropping systems. For examples see: Dively (2005), Sisterson et al. (2004) and Candolfi et al. (2004). Among the species studied are some with combinations of life-history traits (e.g. low reproduction and high emigration) which are sensitive to population declines in areas both inside and outside of IR crop fields (Sisterson et al., 2007).

A prerequisite for such declines is a relatively high mortality in IR fields, compared with non-IR fields (Sisterson et al., 2007). Field studies have shown that some NTAs had lower abundance in IR fields compared with non-IR crop fields, and there are indications that such declines were most likely caused by a shortage of prey caused by control of target pests or a reduction of crop injury that eliminated essential resources; direct toxic effects of IR crops appear the least likely cause. In addition many field studies indicate that currently used IR crops have little or no adverse effects on NTA populations (Reed et al., 2001; Wolt et al., 2001; Al-Deeb & Wilde, 2003; Jasinski et al., 2003; Men et al., 2003; Sisterson et al., 2004; Dively, 2005; Naranjo et al., 2005; Whitehouse et al., 2005; Cattaneo et al., 2006). To date no regional declines in populations of non-target arthropods have been recorded with currently used IR crops. Estimates of effects on populations are generally temporary and within the range of effects from other causes in agricultural environments such as cultivation, crop protection measures and crop varieties. However, some new types of GM PR crops have stacked PR genes which produce more than one toxin or biocide and kill a broader range of targets pests than single products. Stacked IR genes are more likely to reduce NTA populations both inside and outside IR crop fields especially if new GM traits are more toxic to NTAs than current traits. Risk Management Risk assessments using the methodology described above may identify particular sensitive species potentially at risk from the predicted deployment of a GMPR crop. In order to develop

103

appropriate strategies for risk management, a careful analysis of the data compiled during the risk assessment is required. Example 1. If the risk is related to the scale of the cultivation of the GM crop in a region, then the area or frequency of cultivation of the crop can be restricted. This can be done by increasing the refugia areas (i.e. areas with different varieties of the same crop expressing different traits) or preventing continuous cropping with crops expressing the same or related PR genes (Carrier et al., 2004). Example 2. If the risk is associated with effects on field margin and boundary populations of NTAs, then exposure of these populations can be reduced by preventing cultivation of the GM crop close to field boundaries e.g. by establishing non-GM buffer crops around GM crops and close to the boundaries (Caprio et al., 2004). Example 3. If the risk is associated with the synergistic or combined effects of stacked PR genes then these stacks should be restricted as in example 1, but GM varieties expressing single traits could also be deployed in the region. Conclusions Studying effects of PR crops on particular non-targets is a complex and expensive process. Sisterson et al. (2007) propose that life history traits of NTOs such as reproduction and emigration should be among the criteria used to identify species most likely to be affected by IR crops. In addition Snow et al. (2005) and Sisterson et al. (2007) recommended that future risk assessment methods should take into account the regional distribution of IR crops when trying to assess impacts on sensitive NTOs.

This paper suggests that related features such as the presence of non-GMPR crops, other alternative host plants and their distribution in the landscape are also important features for studying.

Risk management strategies need to take full account of the risk analysis and focus on managing the parameters identified as having the greatest effect on NTOs. References Al-Deeb, M.A. & Wilde, G.E. 2003: Effect of Bt corn expressing the Cry3Bb1 toxin for corn

rootworm control on aboveground nontarget arthropods. Environ. Entomol. 32: 1164-1170.

Candolfi, M.P., Brown, K., Grimm, C., Reber, B. & Schmidli, H. 2004: A faunistic approach to assess potential side-effects of genetically modified Bt-corn on non-target arthropods under field conditions. Biocontrol Sci. Technol. 14: 1291-170.

Caprio, M.A., Fave, M.K. & Hankins, G. 2004: Evaluating the impacts of refuge width on source-sink dynamics between transgenic and non-transgenic cotton. J. Insect Sci. 4: 1-5.

Carrière, Y., Dutilleul, P., Ellers-Kirk, C., Pedersen, B., Haller, S., Antilla, L., Dennehy, T.J. & Tabashnik, B.E. 2004: Sources, sinks, and the zone of influence of refuges for managing insect resistance to Bt crops. Ecol. Appl. 14: 1615-1623.

Castellazzi, M.S., Perry, J.N., Colbach, N., Monod, H., Admeczyk, K., Viaud, V. & Conrad, K.F. 2007: New measures and tests of temporal and spatial patterns in agricultural landscapes. Agric., Ecosys. Environ. 118: 339-349.

104

Cattaneo, M.G., Yafuso, C., Schmidt, C., Huang, C., Rahman, M., Olson, C., Ellers-Kirk, C., Orr, B.J., Marsh, S.E., Antilla, L., Dutilleul, P. & Carrière, Y. 2006: Farm-scale evaluation of transgenic cotton impacts on biodiversity, pesticide use, and yield. Proc. Natl. Acad. Sci. U.S.A. 103: 7571-7576.

Dively, G.P. 2005: Impact of transgenic VIP3A X Cry1Ab lepidopteran-resistant field corn on the nontarget arthropod community. Environ. Entomol. 34: 1267-1291.

Ferry, N., Edwards, M.G., Gatehouse, J., Capell, T., Christou, P. & Gatehouse, A.M.R. 2006: Transgenic plants for insect pest control: a forward looking scientific perspective. Transgenic Res. 15: 13-19.

Groot, A.T. & Dicke, M. 2002: Insect-resistant transgenic plants in a multi-trophic context. Plant J. 31: 387-406.

Jasinski, J.R., Eisley, J.B., Young, C.E., Kovach, J. & Willson, H. 2003: Select nontarget arthropod abundance in transgenic and non-transgenic field crops in Ohio. Environ. Entomol. 32: 407-413.

Men, X., Ge, F., Liu, X. & Yardim, E.N. 2003: Diversity of arthropod communities in transgenic Bt cotton and non-transgenic cotton agroecosytems. Environ. Entomol. 32: 270-275.

Naranjo, S.E. 2005: Long-term assessment of the effects of transgenic Bt cotton on the abundance of nontarget arthropod natural enemies. Environ. Entomol. 34: 1193-1210.

Naranjo, S.E., Head, G. & Dively, G.P. 2005: Field studies assessing arthropod nontarget effects in Bt transgenic crops: introduction. Environ. Entomol. 34: 1178-1180.

Reed, G.L., Jensen, A.S., Riebe, J., Head, G. & Duan, J.J. 2001: Transgenic Bt potato and conventional insecticides for Colorado potato beetle management: comparative efficacy and non-target impacts. Entomol. Exp. Appl. 100: 89-100.

Romeis, J., Bartsch, D., Bigler, F., Candolfi, M.P., Gielkens, M.C., Hartley, S.E., Hellmich, R.L., Huesing, J.E., Jepson, P.C., Layton, R., Quemada, H., Raybould, A., Rose, R.I.,Schiemann, J., Sears, M.K., Shelton, A.M., Sweet, J.B, Vaituzis, Z. & Wolt, J.D. 2006: Moving through the tiered and methodological framework for non-target arthropod risk assessment of transgenic insecticidal crops. Proceedings 8th International Symposium on Biosafety of GM crops, Jeju, Korea, pp. 62-67.

Sisterson, M.S., Biggs, R.W., Olson, C., Dennehy, T.J., Carrière, Y. & Tabashnik, B.E. 2004: Arthropod abundance and diversity in Bt and non-Bt cotton fields. Environ. Entomol. 33: 921-929.

Sisterson, M.S., Carrière, Y., Dennehy, T.J. & Tabashnik, B.E. 2007: Nontarget effects of transgenic insecticidal crops: Implications of source-sink population dynamics. Environ. Entomol. 36: 121-127.

Snow, A.A., Andow, D.A., Gepts, P., Hallerman, E.M., Power, A., Tiedje, J.M. & Wolfenbarger, L.L. 2005: Genetically engineered organisms and the environment: current status and recommendations. Ecol. Appl. 15: 377-404.

Whitehouse, M.E.A., Wilson, L.J. & Fitt, G.P. 2005: A comparison of arthropod communities in transgenic Bt and conventional cotton in Australia. Environ. Entomol. 34: 1224-1241.

Wolt, S.J., Burkness, E.C., Hutchison, W.D. & Venette, R.C. 2001: In-field monitoring of beneficial insect populations in transgenic corn expressing a Bacillus thuringiensis toxin. J. Entomol. Sci. 36: 177-187.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 105-110 Ground beetles (Coleoptera: Carabidae) in transgenic herbicide tolerant maize hybrids: Impact of the transgenic crop or the weed control practice? Dóra Szekeres1, Ferenc Kádár2, Zita Dorner1

1Plant Protection Institute, Szent István University, Páter Károly utca 1, H-2100 Gödöllő, Hungary (Email: szekeres.dora@mkk.szie.hu); 2Plant Protection Institute, Hungarian Academy of Sciences, P.O. Box 102, H-1525 Budapest, Hungary Abstract: The cultivation of genetically modified herbicide tolerant (GMHT) maize cultivars and the change in weed control practice by using GMHT crops may have effects on carabid beetles. It is important to know whether the genetical modification, the herbicide regime or a combination of both are responsible for potential effects on carabids. In a field experiment, carabid assemblages from non-glyphosate (commercially available registered herbicide used in maize) treated GMHT maize hybrids with or without insecticidal properties are compared with those of gylphosate treated GMHT ones. Pitfall trapping was conducted during four different growth stages of maize in 2006. Altogether 12335 individuals of 44 carabid species were captured during four weeks. Pseudoophonus rufipes, Poecilus sericeus, Dolichus halensis, Calathus ambiguus and Pseudoophonus calceatus were the most common species in all treatments. The abundance of carabids was significantly higher on non-glyphosate treated maize plots, especially in the second part of the growing season. Recorded weed density was significantly higher in non-glyphosate treated maize plots. We found that the weed control practice applied in GMHT maize is likely to be different from conventional maize crops, which has an impact on the activity density of ground beetles. Key words: genetically modified herbicide tolerant maize, ground beetles, glyphosate, activity density Introduction Ground beetles are often used as indicators in field risk assessments, because the family is rich in species, taxonomically well known, species are abundant in arable fields, and they seem to be very sensitive to habitat changes (Lövei & Sunderland, 1996; Rainio & Niemelä, 2003). In the risk assessment of GM crops, the transgenic cultivar is generally compared to a near-isogenic line. However, in case of GMHT crops, a change in management practice is likely to occur in addition to the genetic modification of the crop. The cultivation of GMHT crops offer more flexible weed control options, because broad spectrum herbicides, like glyphosate, can be applied also post emergence. Glyphosate application causes qualitative and quantitative changes in the weed population (Puricelli & Tuesca, 2005), which can put pressure on the arthropod community. Many carabids (adults and/or larvae) are generalist predators or feed on a mixed diet. The removal of weeds, which are food plants of potential prey species, affects carabids indirectly. Furthermore, carabid larvae are highly sensitive to soil moisture, shadow and temperature, which are influenced by the land coverage on the soil surface. Field studies indicated that the efficacy of weed control practice may have an impact on the abundance and activity density of carabids (Brooks et al., 2005; Lee et al., 2001). We aimed at comparing carabid assemblage structure and activity densities in various genetically modified GMHT hybrids with and without insecticidal properties, treated with glyphosate or with non-glyphosate herbicide (used in common practice).

105

106

Material and Methods Experimental area and sampling method The present work was carried out as part of a larger non-target field study with different transgenic maize hybrids West of Budapest, Sóskút, Hungary in 2006. This paper focuses on the effects on Carabidae of three GMHT maize hybrids: CRHT Coleoptera Resistant and Herbicide Tolerant (glyphosate) (Cry34Ab1, Cry35Ab1xHT; DAS-59122-7xMON-00603-6, Pioneer), LRHT Lepidoptera Resistant and Herbicide Tolerant (Cry1FxHT, DAS-1507-1xMON-00603-6, Pioneer), and CRLRHT Coleoptera Resistant and Lepidoptera Resistant and Herbicide Tolerant (Cry34Ab1, Cry35Ab1xCry1FxHT, DAS-59122-7x DAS-1507-1xMON-00603-6, Pioneer). Plots (25x25 m) were established on chernozem soil and arranged in a randomized complete block design with four replications. Maize was planted with double seeds on 15th-18th May and reduced to 65.000 plants/ha after emergence. Standard agronomic practice typical of the region was used for tillage, fertilization and weed control. The maize plots were treated with commonly used herbicide or with glyphosate. On non-glyphosate treated plots CALLISTO 4SC (active ingredient ai 480g/l mezotrion) and GESAPRIM 500 FW (500g/l atrazine) herbicides were applied at a rate of 144g ai mezotrion/ha and 1000g atrazine ai/ha at the four-leaf stage of maize on 9th-10th June. The glyphosate treated maize plots were treated with ROUNDUP BIOAKTIV herbicide at a rate of 1.06kg ai/ha in broadcast application at the four-leaf (9th June) and the eight-leaf stage of maize (24th June). No insecticides were applied. Plants were harvested on 21th-23th November 2006.

Carabidae (ground beetles) were sampled in each plot with pitfall traps at four different growth stages of maize: mid-vegetative stage (10th-17th June), pre-anthesis (31th July-7th August), main anthesis (07th-14th August) and post anthesis (28th August-4th September). In each pot three pitfall traps were placed at a distance 9 m from each other near the center of each plot and operated for 7 days. Ethylene glycol was used as a killing and preservative agent, and in between sampling dates the traps were filled with soil. Collected carabid adults were identified to species level, using keys by Hůrka (1996).

Weed species and area coverage were estimated on 3x1 m2 on each plot on 13th September when maize plants were still green. Statistical analyses Carabid assemblages of maize plots within the three varieties, treated with glyphosate or non-glyphosate herbicide were compared. The number of species and the biodiversity (Shannon diversity index, Magurran, 2003) were compared using Mann Whitney U-tests. General Linear Models ANOVA (Sokal & Rohlf, 1995) was used to compare the activity density of ground beetles within and across sampling dates. The level of significance was set at p=0.05. Analyses were performed using STATISTICA (Statsoft, 2000). Results and Discussion A total of 12,335 individuals of 44 ground beetle species were collected during the four sampling weeks. The dominant species collected in our sampling area were typical for agricultural fields in Eastern Europe (Lövei & Sárospataki, 1990). Pseudoophonus rufipes (De Geer 1774), Poecilus sericeus Fischer von Waldheim 1823, Dolichus halensis (Schaller 1783), Calathus ambiguus (Paykull 1790), and Pseudoophonus calceatus (Duftschmid 1812) (in order of decreasing activity density) were the most common species both on non-glyphosate and glyphosate treated plots across all maize varieties. No significant differences were found in the number of ground beetle species and the diversity indices between the two

107

treatments for two of the three maize hybrids (Mann Whitney U-test, Table 1). In contrast, in each maize variety the mean number of individuals/trap was significantly higher in non-glyphosate treated plots than in glyphosate treated ones (Mann Withney U-test; Table 1, Figure 1). Table 1. Carabid assemblages in non-glyphosate and glyphosate treated GMHT maize plots. Values (mean±SE) with different letters indicate significant differences (p=0.05) for non-glyphosate and glyphosate treated plots (Mann Whitney U-test). (Sóskút, Hungary, 2006).

CRHT LRHT CRLRHT Treatment/ Characteristics non-gly gly non-gly gly non-gly gly Total no. species 22 25 26 23 27 19 No. species./ trap 9.33±1.67a 9.00±2.04a 10.75±2.14a 9.08±1.56a 11.08±2.91a 8.42±2.19b

MW U-test No. species / trap

Z4=0.328 p=0.743

Z=1.909 p=0.056

Z=2.244 p=0.025

Total no. individuals 2452 1639 2217 1742 2572 1713

No. indiv./ trap 204.33±47.17

a 136.58±36.04

b 184.75±35.22a 145.17±55.42b 214.33±62.52a 142.75±26.39bMW U-test

No. indiv / trap Z=3.002 p=0.003

Z=2.744 p=0.006

Z=2.599 p=0.009

Shannon diversity 0.84±0.20a 0.76±0.18a 0.94±0.20a 0.80±0.19a 0.94±0.17a 0.69±0.21b MW U-test

Shannon diversity Z=1.039 p=0.299

Z=1.443 p=0.149

Z=2.887 p=0.004

Shannon equability 0.38±0.09a 0.35±0.06a 0.39±0.06a 0.36±0.07a 0.39±0.05a 0.32±0.07b MW U-test

Shannon equability Z=0.866 p=0.386

Z=1.270 p=0.204

Z=2.309 p=0.021

Figure 1. Abundance of ground beetle assemblages on non-glyphosate and glyphosate treated GMHT maize plots. Within varieties different letters indicate significant differences, p=0.05.

óskút, Hungary, 2006).

(S

108

When comparing carabid activity densities for each sampling period, significant differences were evident in second part of the maize growing season. The activity density was significantly higher on non-glyphosate treated maize plots at main and post anthesis for each hybrid (GLM; CRHT: F=5.696, d.f.=3, p=0.003; LRHT: F=2.892, d.f.=3, p=0.049; CRLRHT: F=2.894, d.f.=3, p=0.049; Figure 2). Most likely, this finding can be attributed to Pseudoophonus rufipes. This autumn breeding species was the most common one in our study, making up to 79% of the total number of individuals captured.

Figure 2. Seasonal activity patterns of all carabid beetles on non-glyphosate and glyphosate treated LRHT maize plots. Different letters indicate significant differences, p=0.05. (Sóskút, Hungary, 2006).

Figure 3. Percent weed coverage on non-glyphosate and glyphosate treated GMHT maize plots. Within varieties different letters indicate significant differences, p=0.05. (Sóskút, Hungary, 2006).

109

Similar to the abundance of carabids, the weed coverage was significantly higher on non-glyphosate treated maize plots than on glyphosate treated plots in two varieties, but not for CRHT (Mann Withney U-test, CRHT: Z=1.742, p=0.081; LRHT: Z=2.021, p=0.043; CRLRHT: Z=2.309, p=0.021, Table 2). Even though the difference for CRHT was not statistically significant, the weed density was two times higher on non-glyphosate treated maize plots than on glyphosate treated ones (Figure 3). Table 2. Weed species and their percent coverage on non-glyphosate and glyphosate treated GMHT maize plots (+ indicates <1%; Sóskút, Hungary, 2006).

CRHT LRHT CRLRHT Weed species/Treatment non-gly gly non-gly gly non-gly gly Amaranthus retroflexus 4.17 + 0.42 + 2.92 - Convolvulus arvensis 3.33 3.42 3.75 2.58 1.75 2.33 Echinocloa cruss-gali 15.42 2.50 15.83 3.58 16.67 4.00 Hibiscus trionum - - - 0.92 - 0.08 Rubus caesius - 4.25 - - - 0.42 Sorghum halepense 0.83 0.08 1.75 - 3.83 0.08 Total no. of species 5 10 9 5 5 7 Total weed coverage (%) / plot 24.33 12.42 25.17 8.83 27.67 7.92

According to Brooks et al. (2005) the seed-feeding carabids (Pseudoophonus rufipes and Amara spp.) seem to be highly sensitive to weed abundance and appear to be good indicator species for changes in herbicide management on effectiveness at the ecosystem level. Similar to weed coverage and our results of total ground beetle catches, there were significant differences in the abundance of P. rufipes between the treatments for all hybrids (Mann Withney U-test; CRHT: Z=2.946, p=0.003; LRHT: Z=2.512, p=0.012; CRLRHT: Z=2.196, p=0.028, Figure 4).

Figure 4. Abundance of Pseudoophonus rufipes on non-glyphosate and glyphosate treated GMHT maize plots. Within maize varieties different letters indicate significant differences, p=0.05. (Sóskút, Hungary, 2006).

110

In conclusion, we found that the weed control practice using glyphosate applied in GMHT maize, which is different from conventional weed management, reduced the activity density of ground beetles. While conventional weed control practice may vary in efficacy, glyphosate weed control in GMHT maize is likely to be more effective. This may have consequences for ground beetle densities, as shown in our field experiment. For farming practice, this suggests that the choice of a more or less effective weed control and the decision on weed tolerance levels has an impact on carabid densities and not the GMHT crop itself.

The cultivation of GMHT crops depends on several factors, (like agronomic practice, timing of herbicide application, frequency and applied dosage of specific or other herbicides). There are also differences in farming conditions and weed control practices from region to region and season to season (Sweet & Lutman, 2006). For sustainable maize production and weed management, relevant national herbicide application guidelines should be developed for new transgenic herbicide tolerant maize hybrids. Acknowledgements The authors would like to thank Prof. A. C. York (Purdue University, West Lafayette, Indiana) for his constructive comments and suggestions on this manuscript. References Brooks, D.R., Clark, S.J., Perry, J.N., Bohan, D.A., Champion, G.T., Firbank, L.G.,

Haughton, A.J., Hawes, C., Heard, M.S. & Woiwod, I.P. 2005: Invertebrate biodiversity in maize following withdrawal of triazine herbicides. Proc. R. Soc. B. 272: 1497-1502.

Hůrka, K. 1996: Carabidae of the Czech and Slovak Republics. Kabourek, Zlín. 565 pp. Lee, J.C., Menalled, F.D. & Landis, D.A. 2001: Refuge habitats modify impact of insecticide

disturbance on carabid beetle communities. J. Appl. Ecol. 38: 472-483. Lövei, G.L. & Sárospataki, M. 1990: Carabid beetles in agricultural fields in Eastern Europe.

In: Stork, N.E. (ed.), The role ground beetles in ecological and environmental studies. Intercept Ltd., Andover, pp. 87-95.

Lövei, G.L. & Sunderland, K.D. 1996: Ecology and behavior of ground beetles (Coleoptera: Carabidae). Ann. Rev. Entomol. 41: 231-256.

Magurran, A.E. 2003: Measuring ecological diversity. Blackwell, Oxford. 256 pp. Puricelli, E. & Tuesca, D. 2005: Weed density and diversity under glyphosate-resistant crop

sequence. Crop Protection 24: 533-542. Rainio, J. & Niemelä, J. 2003: Ground beetles (Coleoptera: Carabidae) as bioindicators.

Biodiversity and Conservation 12: 487-506. Sokal, R.R. & Rohlf, F.J. 1995: Biometry, 3rd Edition. W. H. Freeman and Company, New

York, 887 pp. StatSoft. 2000: STATISTICA for Windows, I-III. StatSoft Inc. Tulsa, O.K. Sweet, J.B. & Lutman, P.J.W. 2006: A commentary on the bright programme on herbicide

tolerant crops and the implications of the bright and farm scale evaluation programmes for the development of herbicide tolerant crops in Europe. Outlooks on Pest Management 17: 249-254.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 111-115 Reduction of mycotoxin threats to mammals and birds through the cultivation of Bt maize cultivars in Poland Agata Tekiela1, Robert Gabarkiewicz2 1Regional Experimental Station, Institute of Plant Protection, Langiewicza 28, 35-001 Rzeszów, Poland; 2Monsanto Poland, Domaniewska 41, 02-672 Warszawa, Poland (E-mail: robert.gabarkiewicz@monsanto.com) Abstract: The aim of this work was to study Fusarium spp. occurrence and mycotoxins content in grain of four non-transgenic maize cultivars and their genetically modified (Bt) counterparts in the years 2005 and 2006 in the southern part of Poland. Fusarium infestation on corn cobs as well as mycotoxin content in grain was evaluated in the four non-transgenic varieties DKC 3420 (FAO 250), PR39F58 (FAO 270), PR39D81 (FAO 260), PR38F70 (FAO 260), and in their genetically modified counterparts DKC 3421YG, PR39F56, PR39D82, PR38F71. The field trials were located in the Podkarpackie and in the Wielkopolskie provinces in the southern part of Poland. In both years, our field experiments on selected MON 810 maize cultivars showed substantially reduced cob infestations by larvae of Ostrinia nubilalis Hbn. securing a higher yield and lower ear rots infection of kernels, confirming similar results obtained in other countries growing Bt insect-resistant cultivars. The chemical analysis of mycotoxin contamination originating from Fusarium in kernels of Bt maize cultivars grown in southern Poland showed only traces. Key words: mycotoxins, Bt maize cultivars, Poland Introduction The importance of diseases of fungal or bacterial origin, infecting in lower or higher degree all maize plantations, is often underestimated by farmers in Poland (Tekiela, 2001). Fungi of the Fusarium group are considered as the main pathogens attacking maize seedlings, roots, stem bases and cob kernels. During the milky and dough stage of kernels, the cottony, later pink or red mycelium, depending on a fungus species, is apparent on infected, often disintegrated kernels (Lisowicz et al., 2004). Although the Fusarium kernel rot only occasionally cause substantial yield reduction, it affects the quality of kernels used for human consumption or animal feed (Tanaka et al., 1988; Shephard et al., 1996). Other fungi infecting kernels belong to the genera Trichoderma, Penicillium and Trichothecium. Maize ears infested by European corn borer (Ostrinia nubilalis Hbn.) are usually infected by ear rots producing organic compounds such as T-2 toxin, diacetoxyscirpenol, ochratoxin, fumonisins, zearalenone and deoxynivalenol, all toxic to mammals and birds (Nelson et al., 1993; Prelusky et al., 1994). Up to now, monitoring of the content of these toxic mycotoxins in maize kernels was carried out only on an irregular basis. A new directive of the European Union on mycotoxin contamination in agricultural products mandates that attention to this problem shall now not only be given by institutions responsible for food safety and processing industry, but also by farmers.

The aim of this work was to study Fusarium spp. occurrence and mycotoxins content in grain of four non-transgenic maize cultivars and their genetically modified (Bt) counterparts in the years 2005 and 2006 in the southern part of Poland.

111

112

Material and Methods Fusarium infestation on corn cobs as well as mycotoxin content in grain was evaluated in the 4 non-transgenic varieties DKC 3420 (FAO 250), PR39F58 (FAO 270), PR39D81 (FAO 260), PR38F70 (FAO 260), and in their genetically modified counterparts DKC 3421YG, PR39F56, PR39D82, PR38F71. The field trials were located in the Podkarpackie and in the Wielkopolskie provinces in the southern part of Poland.

At harvest, the degree of fungal infection was evaluated on the basis of 100 randomly collected plants (25 plants in 4 replications). Evaluation was performed on the basis of a 5 degree scale (Table 1). Statystical analysis was performed according to Student’s t-test. Table 1. Scale used for the evaluation of cobs infestation by Fusarium spp.

Degree Description 1 2 3 4 5

Very small (1-6 ears, 2%) Small (7-30 ears, 3-10%)

Medium (1/3 of cob, 11-30%) Large (1/2 of cob, 31-50%)

Very large (> 1/2 of cob, 51-100%)

Fusarium species were identified on the basis of conidial morphology at the Rzeszów Regional Experimental Station of the Plant Protection Institute in Poznan.

The analysis of fungal mycotoxins focused on fumonisins (B1, B2, B3) and deoxynivelanol (DON). HPLC method with fluorescent detection was used to determine mycotoxins levels in grain. The analysis was conducted in cooperation with the University of Kazimierz Wielki in Bydgoszcz. The samples were purified on FumoniTest columns from Vicam (fumonisins) and DONtest columns from Vicam, (deoksyniwalenol) according to manufacturer procedures. Results and Discussion As shown in tables 2, 3 and 4, the infestation of fungi was constantly higher on non-transgenic corn cultivars and varied from 21% to 95,35% depending on years. The transgenic counterparts were much less infested. The spectrum of Fusarium species was as follows: Fusarium subglutinans (Wollenw et Reiking) (Nelson et al., 1993), Fusarium graminearum (Schwabe) and Fusarium culmorum (W.G. Smith).

The chemical analysis of mycotoxin contamination originating from Fusarium in kernels of Bt maize cultivars grown in southern Poland showed only traces. Fumonisins analysis indicated that the total concentration in the grain of non-transgenic cultivars ranged from 165 ppb to 511 ppb in 2005 and from 18 ppb to 847 in 2006 (Tables 5, 6). Deoxynivalenol concentration ranged from < 50ppb up to 502 ppb (Table 5). In conclusion, in both years, our field experiments on selected MON 810 maize cultivars showed substantially reduced cob infestations by larvae of Ostrinia nubilalis Hbn. securing a higher yield and lower ear rots infection of kernels. These facts confirmed results from different countries (Hammond et al., 2004; Munkvold et al., 1997; Papst et al., 2005) growing Bt insect-resistant cultivars as well as farmer’s opinions that, in regions with Ostrinia infestation, only the cultivation of Bt maize

113

cultivars may guarantee the competitiveness of Polish producers cultivating maize crops (Tekiela et al., 2005). Table 2. Fusarium infestation on corn cobs in 2005.

No Cultivars Average degree of cob infestation

Percentage of cobs affected

1 2 3 4 5 6 7 8

DKC 3420 DKC 3421YG*

PR39F58 PR39F56* PR38F70 PR38F71* PR39D81 PR39D82*

0,96 0,34 1,11 0,12 0,82 0,29 0,61 0,12

95,35 5,72 79,17 14,26 79,76 3,95 82,02 33,04

LSD (0,05) 0,27 15,07 Table 3. Fusarium infestation on corn cobs in 2006 in Wielkopolska.

No Cultivars Average degree of cob infestation

Percentage of cobs affected

1 2 3 4 5 6 7 8

DKC 3420 DKC 3421YG*

PR39F58 PR39F56* PR39D81 PR39D82* PR38F70 PR38F71*

1,5 1,17 2,9 1,9 0,33 0,29 0,86 0,86

74 56 80 59 24 23 42 23

LSD (0,05) 0,67 22,59 Table 4. Fusarium infestation on corn cobs in 2005 year in Podkarpacie.

No Cultivars Average degree of cob infestation

Percentage of cobs affected

1 2 3 4 5 6 7 8

DKC 3420 DKC 3421YG*

PR39F58 PR39F56* PR39D81 PR39D82* PR38F70 PR38F71*

0,66 0,3 0,95 0,12 0,52 0,08 0,46 0,07

50 27 51 6 21 7 28 2

LSD (0,05) 0,28 15,24

114

Table 5. Mycotoxin content in corn grain in 2005.

Fumonisins (ppb) No Cultivars Deoxynivalenol(DON) (ppb) B1 B2 B3 Total

1 2 3 4 5 6 7 8

DKC 3421YG* DKC 3420 PR38F71* PR38F70

PR39D82* PR39D81 PR39F56* PR39F58

<50 502 <50 148 115 283 <50 502

0 409 25 164 0

121 0

342

0 103 8 45 0 44 0 97

0 14 0 0 0 13 0 13

0 511 33 209 0

165 0

439 Table 6. Fumonisins (ppb) content in corn grain in 2006.

Wielkopolska Podkarpacie No Cultivars B1 B2 Total B1 B2 Total 1 2 3 4 5 6 7 8

DKC 3421YG* DKC 3420 PR38F71* PR38F70

PR39D82* PR39D81 PR39F56* PR39F58

684 3739866 104753 249 716 1212

263 988 203 283 0 55 151 347

947 47271069133053 304 867 1559

0 56 53 212 53 591 0 18

0 0 18 76 22 256 0 0

0 56 71 288 75 847 0 18

References Hammond, B.G., Campbell, K.W., Pilcher, C.D., Degooyer, T.A.,Robinson, A.E., McMillen,

B.L., Spangler, S.M., Riordan, S.G., Rice, L.G. & Richard, J.L. 2004: Lower Fumonisin mycotoxin levels in the grain of Bt corn grown in the United States in 2000-2002. J. Agric. Food Chem. 52: 1390-1397.

Lisowicz, F. & Tekiela, A. 2004: Szkodniki i choroby kukurydzy oraz ich zwalczanie. S. 52-64. W „Technologia Produkcji Kukurydzy”. A. Dubas (red). Wydawnictwo “Wieś Jutra”, 133 ss.

Nelson, P.E., Desjardins, A.E. & Plattner, R.D. 1993: Fumonisins, mycotoxins produced by Fusarium species: biology, chemistry, and significance. Annu. Rev. Phytopathol. 31: 233-252.

Munkvold, G.P. & Desjardins, A.E. 1997: Fumonisins in maize: can we reduce their occurrence. Plant Dis. 81: 556-565.

Prelusky D. B.; Rotter B. A.; Rotter E. G. 1994. Toxicology of mycotoxins. In Mycotoxins in Grain: Compounds Other than Aflatoxin; Trendholm, H. L., Ed.; Eagan Press: St Paul, MN, pp. 359-403.

Papst C., Utz, H.F., Melchinger, A.E., Eder, J., Magg, T., Klein, D. & Bohn, M. 2005: Mycotoxins produced by Fusarium spp. in isogegenic Bt vs. non – Bt maize hybrids under European corn borer pressure. Agron. J. 97: 219-224.

115

Tanaka, T., Hasegawa, A., Yamamoto, S., Lee, U.S., Sugiura, Y. & Ueno, Y. 1988: Worlwide contamination of cereals by the Fusarium mycotoxins nivalenol, deoxynivalenol and zearalenone. 1. survey of 19 countries. J. Agric. Food Chem. 36: 979-983.

Tekiela, A. 2001: Najważniejsze choroby kukurydzy. Kukurydza rośliną przyszłości. Poradnik dla producentów. Agro Serwis: 29-31.

Tekiela, A., Bereś, P. & Grajewski, J. 2005: Wpływ zwalczania chorób i szkodników kukurydzy na zasiedlenie ziarna przez grzyby i zawartość mikotoksyn. Prog. Plant Prot. Res./Postępy w Ochr. Rośl. Vol 45 (2): 1149-1152.

Shephard, G.S., Thiel, P.G., Stockenstrom, S. & Sydenham, E.W. 1996: Worldwide survey of fumonisin contamination of corn and corn-based products. J. AOAC Int. 79: 671-687.

116

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 117-121

Can plants produced from callus culture be used as near-isogenic standards in comparative analyses of transgenic potato clones? Ramona Thieme, Helmut Griess, Thomas Thieme Federal Centre for Breeding Research on Cultivated Plants, Institute of Agricultural Crops, Rudolf-Schick-Platz 3a, D-18190 Groß Lüsewitz, Germany; Am Sportplatz 14, D-18190 Groß Lüsewitz, Germany; BTL Bio-Test Labor GmbH Sagerheide, Birkenallee 19, D-18184 Sagerheide, Germany (E-mail: r.thieme@bafz.de; tt@biotestlab.de) Abstract: For the analysis of the possible impact of transgenic plants on other organisms the choice of a standard plant is essential. In a study of the induction and analysis of somaclonal variation in the agronomic traits of plants from in vitro callus cultures of stem explants, about 13,000 independently regenerated plants (defined as somaclones) of five cultivars and two potato breeding clones were used. The somaclones were planted in a greenhouse and the tubers grown on in the field. During this process the plants were subjected to the multistage selection procedure commonly used in potato breeding. That is, after transferring the in vitro plants to a greenhouse and the first field generation, the weaker plants and those with abnormal leaf shape, colour and poor tuber development, were discarded. Therefore, depending on the genotype only 8 - 22 % of all the somaclones were selected for the second field generation. The haulm growth, maturity type (length of vegetation period) and yield in terms of tuber number, size, shape, eye depth, starch content and starch yield of the somaclones, grown over a total five years and involving three field generations, were assessed and compared with that of the donor genotypes. The frequencies of similar, negatively and positively different somaclones were determined, which characterized the proportion of variants relative to the total number of somaclones produced in vitro and transferred to a greenhouse. In addition, the frequencies of somaclones with deviations among second generation field-grown clones were calculated.

79 - 94 % of the somaclones were different and discarded in the greenhouse generation because of their abnormal appearance. In the second field generation, depending on genotype and trait, 69 - 98 % of the somaclones were indistinguishable from the donor genotype. Up to 20 %, 28 % and 31 % of the somaclones differed in terms of haulm growth, length of the vegetation period and tuber yield per plant, respectively.

Plants of potato cultivars regenerated from callus culture showed differences in agronomic traits. Besides the similar plants there were those that differed in foliage, maturity type, yield and tuber characteristics. Somaclonal variation may affect the phenotypes of potato clones recovered from callus culture, even in the absence of genetic transformation. Thus, clones produced via callus culture cannot be used as near-isogenic standards in comparative analyses of transgenic potato clones. Key words: callus culture, variability, haulm growth, vegetation period, tuber traits, starch, near-isogenic standard, transgenic plants Introduction Somaclonal variation is a term used to describe a phenomenon caused by genetic or epigenetic changes induced during callus formation when plant tissues are cultured in vitro. Typical genetic alterations are in the number and structure of chromosomes, and in DNA sequence; typical epigenetic related events are gene amplification and methylation. After the callus phase the regenerated plants can show genotypic and phenotypic variation. Current techniques used in genetic transformation experiments, involving callus culture and plant

117

118

regeneration, can result in other traits varying in addition to those expected following gene transfer (Heres et al., 2002).

The original aim of this research was to improve one or more of the agronomic traits of potato cultivars or breeding clones prior to their use in potato breeding programmes. A study of the induction and analysis of somaclonal variation in plants from in vitro callus cultures of stem explants was conducted. The influence of somaclonal variation on the haulm growth, maturity type, yield, tuber number, tuber size, tuber shape, eye depth, starch content and starch yield was analysed. This was done to quantify the degree of variability in particular plant properties caused by unplanned changes, which could influence the results when these plants are used in transformation experiments. Material and Methods About 13,000 independently regenerated plants (=somaclones) of five cultivars, one breeding clone and the clone ‘Phyt. St.’ (averaged results for six potato breeding clones with high resistance to Phytopohthtora) were analysed (Table 1). For callus induction stem internodes of explants from in vitro plants were cultivated on solidified agar plates containing Murashige and Skoog’s medium supplemented with 0.2 mg/l naphthalene-acetic acid, 2 mg/l zeatin and 5 mg/l gibberellic acid. After 14 days the primary callus was transferred to a shoot inducing medium containing 2 mg/l zeatin. Plant regeneration occurred after four - eight weeks, shoots were harvested and propagated in vitro. The somaclones were planted in 12 cm pots in a greenhouse, followed by tuber generations grown in the field over a period of five years and three field generations. The plot size in the first field generation was 3 - 4 plants and in the second 6 - 12 plants. Up to 100 control plots per genotype per somaclone population of each donor, were distributed throughout the field. The characteristics of the somaclones originating from in vitro propagated material were assessed and compared with those of the donor cultivars.

The haulm growth (scores: 9 - very strong, 1 - very weak); length of the vegetation period (days from emergence until maturity), tuber yield (g/plant), tuber number (per plant), tuber weight (g per tuber), starch content of tubers and starch yield were determined. The tuber shape and eye depth were scored using a nine-score-scale.

After the transfer of in vitro plants to a greenhouse (n0 - greenhouse generation ), weak plants and those with irregular leaf shape, colour, mosaic-like symptoms, abnormal leaf surface, and very small, few or very long shaped tubers were discarded. The tubers of the remaining somaclones were planted as the first (n1) and the second field generation (n2). During all generations plants were subjected to the multistage selection procedure commonly used in potato breeding.

A two-population comparison of the performance of the somaclones relative to that of the controls was used. The mean value of a specific characteristics and its standard deviation for the control population were calculated and the confidence intervals, µ ± 1.96 s at p = 0.05, were determined for each population of a somaclone. Those somaclones that fell within these intervals are invariant and indistinguishable from the donor genotype.

The percentage of somaclones that differed significantly, negatively or positively, in a statistical sense (Thieme & Griess, 1996, 2005) were determined and presented as variants relative to the total number of somaclones grown on in a greenhouse (n0) and as variants relative to the number of somaclones grown on in the second field generation after a multistage selection (n2).

119

Results and Discussion Trials are required for regulatory purposes if there is any question of a risk of the transgenic plants having an adverse effect on non-target organisms (NTO’s). It has to be decided which plants can be used in such experiments, in the laboratory, semi-field or field conditions. Usually the impact of transgenic plants on NTO’s is compared with that of an isogenic cultivar of the plant (Figure 1). It is argued that the risk is acceptable if the differences in impact (a) are within the variability of the impacts of other cultivars (c and d). It is believed that changes in plant properties (chemical or physical) caused by the insertion of a new gene (pleotropic effects or insertional mutagenesis) are covered by this comparison. If for any reason this isogenic line is produced using callus culture then certain questions need to be asked. Can these plants be used as near-isogenic standards in comparative analyses? Could it be possible that the plant’s properties may be greatly changed by somaclonal variation? In the worst case the true plant property of the non transformed plant (Cv1) is different from that of Cv1 after passing through the callus phase (Iso) and the differences in impact (b) are much greater than the differences between Cv1, Cv2 and Cv3 (b1 and c1). Field trials were used to obtain data on the extent of somaclonal variation.

Res

pons

e of

NTO

's

0

20

40

60

80

100

120

140

Cv1GM

Iso=cv1*

Cv1 Cv2 Cv3

a

b1 bc1 c

d

Figure1. Hypothetical impact of different Cv’s on the response of NTO’s. Cv1GM - genetically modified potato Cv1; Iso=cv1* – Cv1 after passing through the callus phase; Cv1, Cv2, Cv3 – potato cultivars; a – d see text

120

Table 1. Percentage of somaclones that differed in the second field generation in relation to the total number of somaclones grown in a greenhouse (n0) and the number of somaclones of the harvested second field generation (n2) after a multistage selection process based on haulm growth (scores: 9 - very strong, 1 - very weak); length of the vegetation period (days from emergence until maturity), tuber yield (g/plant), tuber number (per plant), tuber weight (g per tuber), starch content (%), starch yield (g per plant), tuber shape (score scale: 3 = round, 5 = oval, 8 = very long) and eye depth (score scale: 1 = very deep, 9 = very shallow). *averaged results for six breeding clones highly resistant to P. infestans.

Astilla Dorisa N St. 119 N Lipsi N Phyt. St.* Karpina Ora n0 n2 n0 n2 n0 n2 n0 n2 n0 n2 n0 n2 n0 n2

Somaclones (n) 643 55 1078 148 2638 404 2904 637 3081 250 540 111 578 68

Traits Percentage of somaclones that differed Haulm growth 93 20 88 16 87 13 79 4 93 9 83 16 89 10Vegetation period 92 7 87 9 88 20 81 14 94 28 81 10 88 2 Tuber yield 94 31 88 16 88 22 82 18 94 22 83 16 90 16Tuber number 94 29 89 22 86 8 80 11 93 21 83 19 90 16Tuber weight 92 11 92 11 87 12 82 20 93 12 84 23 89 7 Starch content 92 9 88 11 87 15 80 8 94 25 83 16 90 14Starch yield 92 11 88 14 88 23 81 16 93 22 82 12 90 13Tuber shape 93 15 86 1 85 4 79 4 93 14 83 15 90 12Eye depth 92 15 87 7 85 6 80 10 92 13 82 15 89 8

In order to improve potato cultivars or breeding clones the variation in specific agronomic traits of the somaclones was analyzed. Because of obvious differences only 8 - 9 % (Phyt. St., cv. Astilla) to 21 - 22 % (cvs. Karpina, Lipsi N) of the somaclones were selected for planting in the second field generation (Table 1). Because of a high incidence of virus infection the majority of the somaclones of the Phyt. St. (breeding clones with increased resistance to late blight) and cv. Ora were discarded. These clones and the cv. are known to be susceptible to viruses. Nevertheless, the results indicate that for the agronomic traits recorded for the somaclones from the seven donors the percentage of plants that were similar, i.e. indistinguishable from the donor genotypes, was only between 6 % (tuber yield and number for cv. Astilla, most traits for Phyt. St.) and 21 % (haulm growth and tuber shape for cv. Lipsi N) of the original number of somaclones. Depending on genotype and trait, 69 - 98 % of the somaclones were indistinguishable from the donor genotype in the second field generation (Table 1). As a consequence, 2 - 31 % of these somaclones differed in the agronomic traits measured (Table 1). The degree of variation depends on the genotype and the specific characteristics. Depending on genotype, up to 20 %, 28 % and 31 % of the somaclones in the second field generation differed, in terms of haulm growth, length of the vegetation period and tuber yield per plant, respectively (Table 1). The most marked difference (29 %) was recorded in cv. Astilla, for tuber number and tuber yield (31 %), in Phyt. St. for length of vegetation period and starch content (28 % and 25 %), respectively (Table 1). In cvs. Karpina and Lipsi, 23 % and 20 %, differed in tuber weight.

121

In general, there were somaclones of all potato donors that differed in haulm growth, vegetation period, yield and tuber traits. These variant potato clones remained phenotypically stable over three generations in the field (Thieme & Griess, 1996, 2005).

On average, of seven genotypes, 14 % and 15.3 % of the somaclones differed in tuber number per plant and in tuber weight, respectively. Of the potato somaclones, 13.6 % and 17.4 % differed in starch content and yield, respectively. The lowest variation, of nearly 7 %, was in tuber shape and eye depth (Thieme & Griess, 2005).

The occurrence of deviations and variation in the percentage of plants that differed after transformation of fourteen potato varieties is also reported by Heeres et al., 2002. In their study 1.4 - 21 % of the transformed greenhouse-grown plants expressed phenotypic differences in leaf shape, leaf colour, vigour, rooting system and tuber development compared to the wild-type cultivar. The total percentage of clones that differed from the donor cultivar in a greenhouse trial was 10.3 %. The differences depended on the genotype. For cv. Calgary and cv. Kardal, 50 % and 9.7 % of the transgenic field-grown plants differed, respectively. On average, 14.3 % of the clones of these cvs were discarded in field trials (Heeres et al., 2002). The authors assume that the differences were due to somaclonal variation induced by the transformation procedure or an inserted gene, and conclude that the use of existing cvs. could lead to transgenic cultivars that are agriculturally inferior to the original cv.

All potato donors produced somaclones, with different degrees of variation in several agronomic traits. These results demonstrate that somaclonal variation may substantially affect the phenotypes of potato clones recovered from callus culture, even in the absence of genetic transformation. In conclusion, clones produced via callus culture cannot be assumed to be near-isogenic when used in comparative analyses of transgenic potato clones. Acknowledgements We are grateful to Tony Dixon for his invaluable help. References Heeres, P., Schippers-Rozenboom, M., Jacobsen, E. & Visser, R.G.F. 2002: Transformation

of a large number of potato varieties: genotype-dependent variation in efficiency and somaclonal variability. Euphytica 124: 13-22.

Thieme, R. & Griess, H. 1996: Somaklonale Variation des Krautes, der Vegetationslänge und des Ertrages bei Kartoffeln. Potato Research 39: 355-365.

Thieme, R. & Griess, H. 2005: Somaclonal variation in tuber traits of potato. Potato Research 48: 153-165.

122

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 123-128 A method for selecting non-target organisms for testing the biosafety of GM plants Jacqui H. Todd, Padmaja Ramankutty, Louise A. Malone Horticulture and Food Research Institute of New Zealand Ltd, Auckland, New Zealand (E-mail: LMalone@hortresearch.co.nz) Abstract: An essential requirement for GM plant risk assessment is the determination of impacts on non-target invertebrates. However, the potential list of non-target invertebrates in any agro-ecosystem is vast, and only a sub-set of these can be tested. We aimed to develop a rational, repeatable method for selecting non-target species for testing the biosafety of GM plants. Protocols for GM biosafety testing suggest a variety of non-target organism selection criteria, all of which are encompassed by the following five questions: could this crop pose a hazard to this organism?; will this organism be exposed to the hazard?; will there be impacts on ecosystems if this organism is affected?; do people value this organism?; can we perform tests with this organism? We combined these criteria using the following equation to give each candidate species a numerical score corresponding to its suitability as a test organism: Species score = where: H = hazard; E = exposure; R = resilience (ability of organism to mitigate the effects of the hazard); S = status of species in ecosystem (species biomass + number of food web links + special ecosystem function); V = anthropocentric value (economic, social, cultural, ethical concerns); T = ease with which tests may be performed. To test the model, we used the example of a hypothetical Bt-pine forest in New Zealand and compiled a database of biological and ecological information on 80 randomly selected invertebrate species found in NZ pine plantations. This was combined with information about a GM Bt-pine plant. The database was composed of answers to 100 questions for each species, and each answer was given a score out of 10. The database was then interrogated to derive values for each of the model’s parameters (H, E, R, S, V, T) for each species. Final species scores were generated using the model, resulting in a ranked list of non-target species with the most suitable candidate for biosafety testing at the top. Key words: screening model, non-target species selection, transgenic crops, risk assessment Introduction In most countries today, genetically modified (GM) plants are subject to laws which require environmental risk assessment before commercial release, and most include provisions for the protection of biodiversity or minimising harm to non-target organisms (e.g. Anon., 1996; EU, 2001). To meet these management objectives, risk assessment methodologies that test specific risk hypotheses and provide measures of potential impacts on non-target organisms (NTOs) are required. While there are a number of laboratory test procedures and techniques available, such as toxicity tests to establish the degree of hazard from a GM plant to a particular organism, it is not practical, or even possible in most cases, to test the responses of all non-target species in the receiving environment. Thus a crucial part of the problem formulation phase of any risk assessment is the selection of non-target test species (Andow et al., 2006; Romeis et al., 2008).

H x E R

x (S + V + T)

123

124

Various criteria for selecting NTOs for biosafety testing of GM plants are currently used by regulators or have been proposed in the literature. Five basic ideas underpin all of these: 1) hazard 2) exposure 3) ecological impacts 4) anthropocentric concerns and 5) ability to perform tests. Information on potential hazard to each NTO can be gathered from toxicity tests with the species itself, related species or surrogate species, and from knowledge of the mode of action of the expressed protein. Exposure can be estimated using information on the spatial and temporal distribution of the NTO, the expression patterns of the new trait in the GM plant, and the feeding habits of the NTO. Estimating ecological impact, i.e. the probability that an impact on a NTO will lead to a significant change in its ecosystem, is far more difficult to estimate and potentially very complex. However, some guidance may be gained from data on the sizes and population densities of NTOs, their food web links, and any specialist ecological roles not directly related to trophic interactions, such as pollination or seed dispersal, even for species for which ecological data are limited. The ability of populations and individual NTOs to be resilient to environmental changes and impacts of a hazard must also be factored in. Anthropocentric concerns are usually reasonably well-documented; for example, rare and endangered species can be found on “red lists” and beneficial species, such as pollinators and natural enemies, are often well-known. The ability to perform tests with a NTO can be estimated using information on the ease with which it can be collected from the field and reared in the laboratory, and the existence of relevant biosafety test protocols.

Given that every crop has a large number of potential NTOs, and applying all aspects of the five criteria above will involve consideration of many characteristics, a comprehensive, transparent and repeatable selection process would be difficult to achieve without the aid of a computer tool. Here we describe a NTO species selection model which interrogates a database of potential NTOs, for suitability based on the five criteria described above, and produces a prioritised list of NTOs for consideration. A more detailed description of this model is given in Todd et al. (2008). Materials and Methods In order to compile information on NTOs relevant to the five criteria identified above, a database was developed using Microsoft® Access 2003, which recorded NTO species biological and ecological attributes as answers to a series of questions. Answers could be entered directly (e.g. “species dry weight”) or chosen from a drop-down menu of options (e.g. “conservation status” = rare, threatened, or common). Space was provided for recording relevant references and other notes. Each attribute either consisted of a numerical value (e.g., dry weight in mg) or was assigned one on the basis of that attribute’s relevance to a set of parameters that defined each criterion. For example, the “value to people” (V) criterion was defined and calculated using the following five parameters: V = V1 + V2 + V3 + V4 + V5 where: V1 = value of the NTO to indigenous people (such as New Zealand Māori) V2 = conservation value V3 = value of the NTO to society in general V4 = economic value V5 = links from the NTO to valued species at higher levels in the food web (e.g., native bird species).

125

The value of each parameter for each species was determined by the scores it received for the attributes relevant to that parameter recorded in the database. For some parameters, there were several attributes, and each of their scores was included in the total. For example, for V1, the attributes from two questions were used: “does this NTO have a name given by the indigenous culture?” and “is this species of value to the indigenous culture?”. Each possible attribute for each question had been assigned a “score” a priori by a subjective (but clearly stated) process based on our expert opinions. To the first question, a score of 10 was given if there was a Māori name, a score of 0 was given if there was a record that no names had ever been given to this insect and in the remainder of cases a score of 5 was given for “unknown”. For the second question, 10 points were awarded if the species had a documented use as a food item or had been mentioned in folklore or otherwise and 5 points were awarded if there were no such records. Thus V1 could have a final value of up to 20 points. In other cases, one attribute could inform more than one parameter, and the scores assigned might be different depending on the parameter being addressed. For example, when the question “is the species rare?” was used to gain information about the species’ value to people, 10 points were awarded if rare or endangered, 5 points if of unknown conservation status, and 0 points if common. However, when using this attribute to gain information about the species’ biomass and its ecosystem status, a rare species scored 1 point, a common species scored 10 and a species of unknown conservation status scored 5 points.

The scoring system also provided a means for dealing with knowledge gaps and uncertainty. We made an a priori decision to give a median score of 5 to any “unknowns”. This ensured that species for which there were little published data were not overlooked, but neither were they unduly emphasised. Other users of the model may have reasons to assign other values to unknowns; these can be recorded and the model will ensure that they are applied across all attributes. For some attributes only very limited information was available for any species producing a source of uncertainty in the data. In these cases, surrogate attributes were used and several of these combined to estimate a parameter. For example, there is little detailed information on the feeding behaviour of many New Zealand insects in relation to particular plants, but information on general feeding habits and mobility of feeding stages can be used to estimate the likelihood of a species being able to shift to an alternative food source in order to avoid a Bt toxin. To indicate the lack of certainty around these attributes a score < 10 but > 0 was assigned.

Compiling information for the “ecological impact” criterion was the most challenging task. In the absence of complex models describing the roles of invertebrates in New Zealand pine forest ecosystems, we chose some simple parameters to give an estimate of each species’ status in the ecosystem (S). These parameters were biomass (estimated using dry weight and population density records), numbers of food web links, and any special ecosystem functions not directly related to trophic interactions, such as pollen or seed dispersal, vectoring of diseases and parasites, and deliberate biological control introductions. Attributes that provided information for these three parameters were summed to give a description of each species’ “ecosystem status” (S). Thus species of high biomass (large and/or numerous invertebrates), with many food web links, and with special ecological functions outside their direct food webs, had high status, meaning that if these species were lost then a significant change in the ecosystem may be triggered.

Because ecosystems are always subject to change and have mechanisms for minimising its impact, we added the concept of resilience to the five criteria in the model. Resilience (R) was defined as the means by which individuals and populations could mitigate the hazard posed by the GM plant. Resilience parameters used in the model included the likelihood of the species developing a genetically-based resistance to the hazard (as in Bt-resistance), any

126

behaviours it may have that would enable it to avoid the hazard (e.g. attributes indicating an ability to feed on other plant species), the potential for populations to migrate to avoid the hazard (including attributes that describe species mobility and ability to survive in other ecosystems such as native forests), and the ability of populations to recover or re-establish after having avoided the hazard (including attributes that described reproductive rate, presence in adjacent ecosystems, and mobility).

The scores for the five criteria and resilience were combined to obtain a Priority Ranking of Non-Target Invertebrates (PRONTI) score for each species:

PRONTI score =

All species in the database were then listed in the order of their PRONTI scores, from

highest to lowest, to identify the priority NTOs for biosafety testing. To test this model, we used the example of a New Zealand forest of GM pine trees

expressing the Bacillus thuringiensis (Bt) toxin, Cry1Ac, and compiled a database of biological and ecological information on 80 invertebrate species found there. We made no attempt to pre-select species in terms of their classification or any other characteristic, but simply chose publications at random and entered data on the first 80 species described in them. To help evaluate the selection procedure, “dummy” species for which specific attributes were assigned were also included. This small test database may not adequately represent the full range of species in New Zealand pine plantations. For example, although it includes several freshwater invertebrates, because many pine forests have streams running through them, some terrestrial groups are not represented. For example, there are no species representing spiders, cicadas, or Orthoptera, although these are most certainly encountered in this agro-ecosystem. More species are currently being added and the resultant larger database will produce more robust results from the model. Results and Discussion Table 1 shows the first 35 of the 80 invertebrate species from the database in order of priority for biosafety testing with a GM pine tree expressing a Cry1Ac Bt toxin. Three “dummy” species are also included: Dummy 1’s attributes were all unknown, so each received a score of 5, Dummy 2 received maximum scores for all criteria, and Dummy 3 received a zero score for the hazard criterion. These scores dictate that the dummy species will be in the middle, top and bottom of the list respectively. A larger database, including more invertebrate species that better represent all the groups found in this agro-ecosystem, would obviously yield a different and more comprehensive list. Such a database is presently being constructed so that the model can be used for applications in both New Zealand pine plantations and pastures.

The priority list is not designed to be used as a definitive list of test species, but as a resource for biotechnologists and regulators to discuss priorities and optimise allocation of resources for pre-release GM plant biosafety testing. Because the model is designed to be useful for applications beyond today’s pest resistant GM crops and in many different agro-ecosystems, we have not made a preliminary distinction between “target” and “non-target” organisms. Thus some of the species near the top of the list could well be targets for the GM plant, and for these the biotechnologist will already have ample data on effects. Other species, for which the potential ecological impacts might be great, may not have been obvious candidates for biosafety testing prior to running the model. For example, freshwater insects

H x E R

x (S + V + T)

127

with adult stages that feed on nectar may be important pollinators of some native plants in the forest, even though such insects do not readily come to mind in this respect. Their appearance near the top of the list may prompt us to formulate some testable hypotheses and conduct experiments to determine whether the concerns the model has raised are justified. New findings can then be added to the database and the model re-run to produce a new list based on better information. Table 1. Priority list of invertebrate species for testing the biosafety of Bt (Cry1Ac)-expressing GM pine for growing in New Zealand (35 species of 80 shown). Note that the PRONTI score has no ecological meaning and is useful only for ranking the invertebrates relative to each other. Criteria Scores Rank Species Name Order H E R S T V PRONTI

1 Dummy 2 Lepidopt. 20 40 209 96 80 80 982 2 Ctenopseustis obliquana Lepidopt. 10 39 147 97 80 63 635 3 Epiphyas postvittana Lepidopt. 10 39 160 121 80 33 568 4 Pseudocoremia suavis Lepidopt. 9 40 157 107 80 42 526 5 Planotortrix notophaea Lepidopt. 10 39 152 54 78 63 498 6 Austroperla cyrene Plecopt. 13 40 168 60 50 48 482

7 Helicoverpa armigera conferta Lepidopt. 10 39 165 85 80 28 456

8 Aoteapsyche species Trichopt. 10 36 171 89 70 58 449 9 Cnephasia jactatana Lepidopt. 7 40 152 85 80 63 420

10 Hydrobiosis species Trichopt. 13 36 179 70 40 48 408 11 Ctenognathus cardiophous Coleopt. 10 27 95 51 46 40 404 12 Hydrobiosis centralis Trichopt. 13 36 168 56 40 48 394 13 Ctenognathus crenatus Coleopt. 10 27 97 57 39 45 394 14 Triplectides obsoletus Trichopt. 11 36 160 55 50 48 379 15 Zelandoperla fenestrata Plecopt. 11 40 127 32 39 38 372 16 Psilochorema species Trichopt. 13 36 181 60 40 45 369 17 Aulacopodus calathoides Coleopt. 10 27 102 49 45 45 368 18 Zelandobius furcillatus Plecopt. 11 36 168 57 59 38 363 19 Ctenognathus novaezelandiae Coleopt. 10 27 110 53 55 40 363 20 Ctenognathus bidens Coleopt. 10 27 112 56 46 40 340 21 Costachorema xanthopterum Trichopt. 13 36 170 47 30 48 338 22 Meteorus pulchricornis Hymenopt. 10 27 130 66 80 15 336 23 Hydrobiosidae family Trichopt. 13 36 168 43 40 35 328 24 Orthopsyche fimbriata Trichopt. 10 36 166 66 40 45 328 25 Hudsonema alienum Trichopt. 11 36 153 43 50 35 326 26 Planotortrix excessana Lepidopt. 6 39 154 70 78 63 319 27 Dummy 1 Araneae 10 20 112 86 40 50 314 28 Neurochorema species Trichopt. 11 36 178 58 40 45 314 29 Holcaspis brevicula Coleopt. 10 27 75 10 36 35 290 30 Acroperla trivacuata Plecopt. 11 36 182 47 39 38 271 31 Antipodochlora braueri Odonata 11 27 136 29 35 58 267 32 Rhyssa persuasoria Hymenopt. 12 27 134 61 35 13 263 33 Liothula species Lepidopt. 6 39 133 70 25 42 240 34 Polyplectropus species Trichopt. 10 27 168 52 50 48 240 35 Declana leptomera Lepidopt. 6 39 135 41 45 47 231

128

Further studies are planned with scenarios where GM plants have already been tested and released, to compare the priority NTOs generated by the model with those used in the original regulator’s assessment of the crop, e.g. Bt-corn in a European situation.

The model offers a flexible, logical and transparent way of applying selection criteria to a set of organisms. It can easily be adapted for use in other agro-ecosystems and with other kinds of plant modifications. If the user considers that some criteria are more important than others, then the ranking system can be altered to reflect this. It is also possible to expose significant ecological knowledge gaps by assigning a higher rank to the “unknown” answers, thus identifying potential limitations to our ability to conduct the best possible biosafety tests with GM plants. Acknowledgements We thank our colleagues at HortResearch who contributed ideas to this project, particularly E. Barraclough, B. Philip, E. Burgess, N. Markwick and J. Poulton, and C. Walter of Scion, New Zealand, for provision of Bt GM pine trees. This work was funded by New Zealand’s Foundation for Research Science and Technology, contract C06X0222. References Andow, D.A., Birch, A.N.E., Dusi, A.N., Fontes, E.M.G., Hilbeck, A., Lang, A., Lovei, G.L.,

Pires, C.S.S., Sujii, E.R., Underwood, E. & Wheatley, R.E. 2006: Non-target and biodiversity risk assessment for genetically modified (GM) crops, 9th International Symposium on the Biosafety of Genetically Modified Organisms, September 24-29, 2006, pp. 68-73.

Anon. 1996: Hazardous Substances and New Organisms Act 1996. New Zealand Government, Wellington, New Zealand.

EU 2001: Directive 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC, http://ec.europa.eu/environment/biotechnology/pdf/dir2001_18.pdf (accessed on 18/10/06).

Romeis, J., Bartsch, D., Bigler, F., Candolfi, M.P., Gielkens, M.M.C., Hartley, S.E., Hellmich, R.L., Huesing, J.E., Jepson, P.C., Layton, R., Quemada, H., Raybould, A., Rose, R.I., Schiemann, J., Sears, M.K., Shelton, A.M., Sweet, J., Vaituzis, Z. & Wolt, J.D. 2008: Assessment of risk of insect-resistant transgenic crops to nontarget arthropods. Nat. Biotech. 26: 203-208.

Todd, J.T., Ramankutty, P., Barraclough, E.I. 6 Malone, L.A. 2008: A screening method for prioritizing non-target invertebrates for improved biosafety testing of transgenic crops. Environ. Biosafety Res. DOI: 10.1051/ebr:2008003. Available online at: www.ebr-journal.org

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 129-132

Impact of transgenic Bt corn on European corn borer (Ostrinia nubilalis Hübner) in Lower Silesia, Poland. Preliminary results J.P. Twardowski1, M. Hurej1, L. Kordas2 1Wroclaw University of Environmental and Life Sciences, Departament of Plant Protection, pl. Grunwaldzki 24a, 50-363 Wroclaw, Poland (E-mail: jacek.twardowski@up.wroc.pl); 2Wroclaw University of Environmental and Life Sciences,Departament of Soil Management and Plant Cultivation, Poland Abstract: The aim of this preliminary study performed in Lower Silesia, Poland, was to determine the impact of transgenic Bt corn (MON 810) on its target pest, the larvae of the European corn borer (ECB), as compared to a non-transgenic isoline control. Abundance of the pest as well as non-target organisms (data not included here) was monitored once a week, from the beginning of moth flight until the end of the corn growing season. At the end of the season, 100 plants from each treatment were taken to the laboratory for further analysis. Apparently higher level of damage caused by ECB larvae was recorded on conventional corn stalks and cobs when compared to transgenic Bt-corn. The differences were noticed throughout the whole growing season. Positive effects of the transgenic cultivar in comparison to non-Bt corn on different parameters of plant damage caused by ECB larvae were confirmed by the laboratory analysis. Key words: European corn borer, corn, Bt corn, non-Bt Introduction In 2006, corn was grown in Poland on approximately 700 thousand hectares, mainly for grain use. It is supposed that the corn cultivation area will increase in the foreseeable future, the main reason for this being growing interests in ethanol production. This situation could cause a greater corn pests incidence as well. In Lower Silesia, the European corn borer (ECB) Ostrinia nubilalis Hübner (Lepidoptera: Crambidae) is the major lepidopteran pest of corn (Zea mays (L.) requiring chemical plant protection. Lately, corn damage caused by larvae has locally increased to 50-80% or even 100% per field. Losses of grain yield were from 20% to 36% (Lisowicz, 2003). Similar yield losses occur in the United States and in some Central European countries (Bohn et al., 2000, Saeglitz et al., 2006; Manachini, 2006). At least two insecticide treatments are recommended for the chemical control of the ECB (Bereś 2006). Given that, at the proper time of chemical application, plants are very high and an efficient and appropriate chemical intervention causes huge inconveniences, the use of transgenic Bt corn expressing the δ-endotoxin Cry 1Ab from Bacillus thuringensis Berliner var. kurstaki could be a convenient method to protect plants against O. nubilalis throughout the whole growing season.

In Poland, until now, there is no information concerning the impacts of Bt corn on organisms occurring on this plant. Therefore, the aim of our three-year study was to determine the impact of Bt corn on target ECB larvae and on non-target arthropods. In the present preliminary study, only the effect of Bt corn on ECB is presented.

129

130

Material and Methods The preliminary study was conducted on one corn field near Wrocław, in Lower Silesia, Poland, in 2006. The field of 0.4 ha was divided into two equal parts. The first part was planted with transgenic Bt corn (MON 810, Monsanto Company, DKc 3421YG cultivar), and on the second part, the corresponding non-transgenic isoline corn (DKc 3420 cultivar) was used. The abundance of pests was monitored once a week, starting on the 20th of June (beginning of moth flight) until the end of the corn growing season. The observations were made along a field diagonal at five sites within each treatment (transgenic and conventional) on ten consecutive plants (50 plants in each treatment in total). The visible plant damage by ECB was recorded on different parts of the plant (stalk, cob). At the end of the growing season, 100 plants from each treatment (ten consecutive plants taken from ten randomly chosen sites) were taken to the laboratory for further analysis. Laboratory analysis consisted in recording the length of the tunnels and the number of holes bored by the larvae. In addition, presence of Fusarium spp. on corn cobs was noted. No insecticide treatments were applied during the growth of the crop. Statistical analysis of the data was performed using one way analysis of variance (ANOVA), and means were separated using Tukey’s HSD (honest significant difference) test. Differences among means were compared at the 5% level of significance.

Results and Discussion During our study, the damage on corn stalks and cobs caused by European corn borer larvae was recorded on clearly higher level in conventional plants when compared to transgenic plants (Figures 1, 2). The differences were noted throughout the whole growing season. Numbers of ECB larvae on corn cobs were significantly higher on non-Bt plants (F = 22.81, df = 1, P = 0.000006). Likewise, significant differences in larvae abundance between conventional and Bt-cultivar were recorded on corn stalks (F = 9.61, df = 1, P = 0.0025). Tukey’s HSD test showed significant differences on corn stalks at four sampling dates at the end of the growing season (Figure 2).

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

20.07 28. 4.08 11. 18 25. 1.09

observation date

mea

n no

. of l

arva

e

Bt non-Bt

Figure 1. Mean number of ECB larvae recorded on corn cobs (per 50 plants).

131

bb

bb

a

a

a

a*

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

20.07 28. 4.08 11. 18 25. 1.09

observation date

mea

n no

. of l

arva

e

Bt non-Bt

Figure 2. Mean number of ECB larvae recorded on corn stalks (* different letters indicate significant difference between treatments (per 50 plants) (Tukey HSD test, p≤ 0.05).

Similarly, positive effects of transgenic cultivar in comparison to non-Bt corn on different parameters of plant damage caused by larvae were found in the lab-analysis (Table 1). Stalk injuries were recorded on 55% of conventional plants and only on 3% of transgenic ones. Similar results were achieved in the case of injured cobs, i.e. 14% and 1%, respectively. The number of tunnels bored by ECB larvae corresponded to 48 (mean length 3.5 cm) in conventional cultivar and only 3 (mean length 0.2 cm) in transgenic cultivar. Both, in respect to the tunnel number and to their length, significant differences were found between the tested corn cultivars (F = 5.66, df = 1, P = 0.019). Almost the same proportion of cobs was infected by Fusarium spp. on Bt and conventional cultivar. Table 1. Injuries of 100 corn stalks and cobs determined in laboratory analysis (* different small letters indicate a significant difference between treatments (ANOVA, p≤ 0.05).

Non-Bt Bt Stalk injuries (%) 55 3

Corn cob injuries (%) 14 1 No. of tunnels in stalk 48 a* 3 b

Mean length of tunnels (cm) 3.5 a 0.2 b Mean no. of larvae 10 0

Fusarium spp. infections of the cobs (%)

50 46

Our results suggest that Cry1Ab expression, in general, has a clear impact on ECB

larvae. Both observations during the growing season as well as the analysis in the laboratory show significantly less density and damage caused by Ostrinia nubilalis larvae in transgenic

132

Bt corn. Similar results were found by many researchers in different countries. In France, for example, Bourget et al. (2002) found much more ECB larvae per stalk on non-Bt corn. The numbers ranged from 5.25 (non-Bt) compared to 1.83 (Bt) at one of the chosen sites. In the U.S., Venditti and Steffey (2002) confirmed the effectiveness of Bt-corn in controlling the target pest. They found 115 larvae per 100 of conventional plants but only 9.6 of Bt corn.

References Bereś, P. 2006: Effects of chemical control of the European corn borer (Ostrinia nubilalis

Hbn.) in south-eastern Poland in 2003-2005. Progress in Plant Protection 43: 464-467. Bohn, M., Schulz, B., Kreps, R., Klein, D. & Melchinger, A.E. 2000: QTL mapping for

resistance against the European corn borer (Ostrinia nubilalis H.) in early maturing European dent germplasm. Theor. Appl. Genet. 101: 907-917.

Bourget, D., Chaufaux, J., Micoud, A., Delos, M., Naibo, B., Bombarde, F., Marque, G., Eychenne, N. & Pagliari, C. 2002: Ostrinia nubilalis parasitism and the field abundance of non-target insects in transgenic Bacillus thuringiensis corn (Zea mays). Environ. Biosafety Res. 1: 49-60.

Lisowicz, F. 2003: Increasing harmfulness of European corn borer (Ostrinia nubilalis Hbn.) on maize in south-eastern Poland. Progress in Plant Protection 43: 247-250.

Manachini, B. 2006: Resistant management of Bt corn and sustainability in Italy. J. Verbr. Lebensm. 1: 109-110.

Saeglitz, C., Bartsch, D., Eber, S., Gathmann, A., Priesnitz, K.U. & Schuphan, I. 2006: Monitoring the Cry1Ab susceptibility of European corn borer in Germany. J. Econ. Entomol. 99: 1768-1773.

Venditti, M.E. & Steffey, K.L. 2002: Field effects of Bt corn on the impact of parasitoids and pathogens on European corn borer in Illinois. In Proceedings of the 1st International Symposium of Biological Control of Arthropods, U.S. Forest Service, Amherst, MA, pp. 278-283.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 133-137 Non-target organism risk assessment in Bt crops Zigfridas Vaituzis United States Environmental Protection Agency, Office of Pesticide Programs, Biopesticides and Pollution Prevention Division, 1200 Pennsylvania Ave. NW, (7511P), Washington, DC 20460.0001 (E-mail: vaituzis.zigfridas@epa.gov) Abstract: This is a concise summary of the USEPA GM plant (PIP) non-target organism risk assessment process used for the last 12 years with citations of the supporting documentation. The summary is meant to frame the issues in a “Regulatory Risk Assessment Science” context since the hazard/risk issue is frequently overlooked when viewed from a pure toxicity perspective. It should be useful to anyone who may want to articulate how the environmental risk assessments were performed on the GM crops currently registered by the USEPA, and to understand that reports of adverse effects to non-target organisms per se do not automatically imply risk under field use conditions. To minimize data requirements and avoid unnecessary tests, the EPA risk assessments are structured such that risk is determined first from estimates of hazard under “worst-case” exposure conditions. A lack of adverse effects under these conditions provides sufficient confidence that there is no risk and no further data would be needed. Such screening tests conducted early in an investigation are broad in scope but relatively simple in design, and can be used to demonstrate acceptable risk in most circumstances. When screening studies conducted in a laboratory setting suggest potentially unacceptable risk, additional studies are designed to assess risk under more realistic field exposure conditions. These later tests are more complex than earlier screening studies. Use of this “tiered” testing framework saves valuable time and resources by organizing the studies in a manner that eliminates unnecessary lines of investigation. The initial, lower tier, high dose screening studies also allow tighter control over experimental variables and exposure conditions, resulting in a greater ability to produce statistically reliable results at relatively low cost. Key words: Bt crops, environmental risk assessment Risk Assessment Process Tiered testing is designed to first represent unrealistic worst case scenarios and ONLY progress to real world field scenarios if the earlier tiered tests fail to indicate adequate certainty of acceptable risk. Screening (Tier I) non-target organism hazard tests are conducted at pesticidal substance concentrations several times higher than the highest concentrations expected to occur under realistic field exposure scenarios. This has allowed an endpoint of 50% mortality to be used as a trigger for additional higher-tier testing. Less than 50% mortality under these conditions of extreme exposure suggest that population effects are likely to be negligible given realistic field exposure scenarios.

The USEPA uses a tiered (Tiers I-IV) testing system to assess the toxicity of a Plant Incorporated Protectant (PIP) to representative non-target organisms that could be exposed to the toxin in the field environment. Tier I high dose studies reflect a screening approach to testing designed to maximize any toxic effects of the test substance on the test (non-target) organism. The screening tests evaluate single species in a laboratory setting with mortality as the end point. Tiers II – IV generally encompass definitive hazard level determinations, longer term greenhouse or field testing, and are implemented only when unacceptable effects are seen at the Tier I screening level. (Field testing was, however, required for the first 10 years

133

134

of PIP cultivation as a precaution due to the uncertainties inherent in the implementation of a new, untried technology.)

Testing methods which utilize the tiered approach were last published by the EPA as Harmonized OPPTS Test Guidelines, Series 850 and 885 (EPA 712-C-96-280, February 1996)i. The Series 885 guidelines apply to microbes and microbial toxins when used as pesticides (as defined in 40 CFR 152.20), including those that are naturally occurring, and those that are strain-improved, either by natural selection or by deliberate genetic manipulation. Therefore PIPs that are microbial toxins are also covered by these testing guidelines.

The Tier I screening maximum hazard dose (MHD) approach to environmental risk assessment is based on some factor (whenever possible >10) times the maximum amount of active ingredient expected to be available to terrestrial and aquatic non-target organisms in the environment (the Estimated Environmental Concentration - EEC)ii. Tier I tests serve to identify potential hazards and are conducted in the laboratory on representative species at high dose levels which increase the statistical power to test the hypothesesiii. Elevated test doses, therefore, add certainty to the assessment, and such tests can be well standardized. The OPPTS Test Guidelines call for initial screening testing of a single group or several groups of test animals at the maximum hazard dose level. The Guidelines call for testing of one treatment group of at least 30 animals or three groups of 10 test animals at the screening test concentration. The Guidelines further state that the duration of all Tier I tests should be approximately 30 days. Some test species, notably non-target insects, may be difficult to culture and the suggested test duration has been adjusted accordingly. In cases where an insect species cannot be cultured for 30 days, the observation period is until negative control mortality rises above 20 percent or until pupation/maturation takes place.

Failing the Tier I (10 X EEC) screening does not necessarily indicate the presence of an unacceptable risk in the field but it triggers the need for additional testing.iv A less than 50% mortality effect at the MHD is taken to indicate minimal risk. However, greater than 50% mortality does not necessarily indicate the existence of unacceptable risk in the field, but it does trigger the need to collect additional dose-response information and a refinement of the exposure estimation before deciding if the risk is acceptable or unacceptable. Where potential hazards are detected in Tier I testing (i.e. mortality is greater than 50%), additional information at lower test doses is required which can serve to confirm whether any effect might still be detected at more realistic field [1X EEC] concentrations and routes of exposurev.

When screening tests indicate a need for additional data, the OPPTS Harmonized Guidelines call for testing at incrementally lower doses in order to establish a definitive LD50 and to quantify the hazard. In the definitive testing, the number of doses and test organisms evaluated must be sufficient to determine an LD50 value and, when necessary, the Lowest Observed Adverse Effect Concentration (LOAEC), No Observed Adverse Effect Level (NOAEL) , or reproductive and behavioral effects such as feeding inhibition, weight loss, etc. In the final analysis, a risk assessment is made by comparing the LOAEC to the EEC. When the EEC is lower than the LOAEC, a no risk determination is made. These ‘definitive’ tests offer greater environmental realism, but they may have lower statistical power. Appropriate statistical methods, and appropriate statistical power, must be employed to evaluate the data from the definitive tests. Higher levels of replication, test species numbers or repetition are needed to enhance statistical power in these circumstances.

Data that shows less than 50 % mortality at the maximum hazard dosage level – (i.e. LC50, ED50, or LD50 >10 X EEC) are sufficient to evaluate adverse effects, making lower, field exposure dose definitive testing unnecessary. It is also notable that the recommended

135

>10X EEC maximum hazard dose level is a highly conservative factor. The published EPA Level of Concern [LOC] is 50% mortality at 5X EEC for terrestrial organismsvi.

Risk Assessment Process ‘Validation’ The tiered hazard assessment approach was developed for the EPA by the American Institute of Biological Sciences (AIBS) and confirmed, in 1996, as an acceptable method for environmental risk assessment by a FIFRA Scientific Advisory Panel (SAP) on microbial pesticides and microbial toxins. The December 9, 1999 SAP agreed that the tiered approach was suitable for use with Plant Incorporated Protectants (PIP); however, this panel recommended that for PIPs with limited host range insecticidal properties, such as Bt Cry proteins, the testing should be expanded to include additional beneficial invertebrates closely related to target species and/or likely to be present in GM crop fields. Testing of Bt Cry proteins on species not closely related to the target insect pest was not recommended, although it is still performed to fulfill the published EPA non-target species data requirements. In addition, the EPA required field testing (Tier IV) for the first 10 years of PIP cultivation to evaluate population-level effects on non-target organisms. The October 2000 SAP, the August 2002 SAP, and some public comments generally agreed with this approach, with the additional recommendation that indicator organisms should be selected on the basis of potential for field exposure to the subject protein.vii This approach was also confirmed the National Academy of Sciencesviii (NAS, 2000). Chronic Studies Proteins do not bioaccumulate. The biological nature of proteins makes them readily susceptible to metabolic, microbial, and abiotic degradation once they are ingested or excreted into the environment. Although there are reports that some proteins (Bt Cry proteins) bind to soil particles at low pH values, it has also been shown that these proteins are degraded rapidly by soil microbial flora upon elution from soil particles. Therefore, since delayed adverse effects and/or accumulation of toxins through the food chain are not expected to result from exposure to proteins, protein toxins are not routinely tested for chronic effects on non-target organisms. The 30 day test duration requirement does, however, amount to subchronic testing when performed at field exposure test doses. In addition, several of the invertebrates routinely tested undergo a developmental and reproductive cycle during the 30 day test period. Conclusions The tiered approach to testing of non-target organisms ensures, to the greatest extent possible, that the Agency requires the minimum amount of data needed to make scientifically sound regulatory decisions. The EPA believes that maximum hazard dose Tier I screening testing presents a reasonable approach for evaluating risks related to the use of plant incorporated pesticidal substances and for identifying negative results with a high degree of confidence. The Agency expects that Tier 1 testing is sufficient to support non-target organism risk assessment for PIP registrations. Field testing (Tier IV) was, however, required for the first 10 years of PIP cultivation as a precaution due to the uncertainties inherent in the implementation of a new, untried technology. No unexpected adverse effects on agricultural field biodiversity were seen.

136

Footnotes

i http://www.epa.gov/opptsfrs/publications/OPPTS_Harmonized/885_Microbial_Pesticide_Test_Guidelines/Series/ ii The dose margin can be less than 10x where uncertainty in the system is low or where high concentrations of test material are not possible to achieve due to test organism feeding habits or other factors. High dose testing also may not be necessary where many species are tested or tests are very sensitive, although the test concentration used must exceed 1X EEC. iii A proportions test provides a statistical approach to evaluate whether an observed proportion of responding individuals is different from a hypothetical proportion. Non-target hazard screening tests that are conducted at exposure concentrations several times higher than the maximum concentrations expected to occur under realistic field exposure scenarios have customarily allowed an endpoint of 50% mortality to be used as a trigger for additional higher-tier testing. Lower levels of mortality under these conditions of extreme exposure suggest that population effects are likely to be negligible given realistic exposure scenarios. Thus, it follows that the observed proportion of responding individuals can be compared to a 50% effect to determine if the observed proportion is significantly lower than 50%. For example, using a binomial approach, a sample size of 30 individuals is sufficient to allow a treatment effect of 30% to be differentiated from a 50% effect with 95% confidence using a one-sided Z test. A one-sided test is appropriate because only effects of less than 50% indicate that further experiments are not needed to evaluate risk. (http://www.epa.gov/pesticides/biopesticides/pips/non-target-arthropods.pdf) iv It is notable that the 10 X EEC MHD testing approach is not equivalent to what is commonly known as “testing at a 10X SAFETY FACTOR” where any adverse effect is considered significant. Tier I screen testing is not ‘safety factor testing’. In a “10X safety factor” test any adverse effect noted is a “level of concern”, whereas in the EPA environmental risk assessment scenario any adverse effect is viewed as a concern only at 1X the field exposure. v The 1X EEC test dose is based on plant tissue content and is considered a high worst case dose (sometimes referred to as HEEC). This 1X EEC is still much greater than any amount which any given non-target organism may be ingesting in the field because most non-target organisms do not ingest plant tissue. vi Environmental Protection Agency (USEPA) (1998). “Guidelines for Ecological Risk Assessment.” EPA 630/R-95-002F. Washington, DC, USA. [Federal Register, May 14, 1998. 63(93): 26846-26924.] The established peer and EPA Science Board reviewed guidance on screening test Levels of Concern (LOC) is 50% mortality at 5X environmental concentration for terrestrial and 10X for aquatic species. The appropriate endpoints in high dose limit/screening testing are based on mortality of the treated, as compared to the untreated (control) non-target organisms. A single group of 30 test animals may be tested at the maximum hazard dose. vii EPA-SAP. February 4, 2000. Characterization and non-target organism data requirements for protein plant-pesticides. SAP report No. 99-06A for FIFRA Scientific Advisory Panel Meeting held December 8, 1999, held at the Sheraton Crystal City Hotel, Arlington, VA.

137

EPA-SAP. March 12, 2001. Bt plant-pesticides risk benefit assessments. SAP report No. 2000-07 for FIFRA Scientific Advisory Panel Meeting held October 18-20 at the Marriott Crystal City Hotel, Arlington, VA. EPA-SAP. November 6, 2002. Corn rootworm plant-incorporated protectant insect resistance management and non-target insect issues. Transmittal of meeting minutes of the FIFRA Scientific Advisory Panel Meeting held August 27-29 at the Marriott Crystal City Hotel, Arlington, VA. EPA-SAP. August 19, 2004. Product characterization, human health risk, ecological risk, and insect resistance management for Bacillus thuringiensis (Bt) cotton products. Transmittal of meeting minutes of the FIFRA Scientific Advisory Panel Meeting held June 8-10 at the Holiday Inn Ballston, Arlington, VA. viii Environmental Effects of Transgenic Plants: The Scope and Adequacy of Regulation is available from the National Academy Press, 2101 Constitution Avenue, N.W., Lockbox 285, Washington, DC 20055; (800) 624-6242 or (202) 334-3313 (in the Washington metropolitan area); http://www.nap.edu.

138

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 139-144

Transgenic Escherichia coli co-expressing cry1Ca and cry1Ac: toxicity and synergy against three agricultural pests Arieh Zaritsky, Eitan Ben-Dov Department of Life Sciences, Ben-Gurion University of the Negev, P.O. Box 653, Be'er-Sheva 84105, Israel (E-mail: ariehz@bgu.ac.il, etn@bgu.ac.il) Absract: The genes cry1Ac and cry1Ca from Bacillus thuringiensis subsps. kurstaki HD-73 and aizawai 4J4, respectively, encoding δ-endotoxins against lepidopteran larvae, were isolated, cloned and expressed in Escherichia coli, separately and together, under control of the early T7, PA1 inducible promoter. Toxicities were examined against larvae of three major agricultural pests: Pectinophora gossypiella, Helicoverpa armigera and Spodoptera littoralis. The clone expressing cry1Ac (pBt-1A) was the most toxic to P. gossypiella (LC50 of 0.27×108 cells g-1). Clone pBt-1CA expressing both genes displayed the highest toxicity (LC50 of 0.12×108 cells ml-1) against S. littoralis, with a synergy factor of 164 between them. Key words: Bacillus thuringiensis, δ-endotoxin, transgenic Escherichia coli, agriculture pests, synergy Introduction Various subspecies of the entomopathogenic bacterium Bacillus thuringiensis (Bt) are specific and hence environmentally safe microbial control agents against larvae of lepidopteran, dipteran, and coleopteran species (Schnepf et al., 1998). The larvicidal activity is included in proteinaceous crystalline δ-endotoxins synthesized during sporulation. Extensive use of Bt spray products to control lepidopteran pests has resulted in a number of field populations with different levels of resistance to the insecticidal crystal proteins (ICPs) (Sayyed et al., 2000; Tabashnik, 1994). Some laboratory-selected insects resistant to Cry1 toxins display cross-resistance to other ICPs; e.g. Pectinophora gossypiella (Tabashnik et al., 2000), Plutella xylostella (Tabashnik et al., 1994), Heliothis virescence (Gould et al., 1995) and Spodoptera exigua (Moar et al., 1995). Various strategies have been suggested to avoid selection for resistance in these pests, including refuges of susceptible plants along with transgenic insect-resistant plants and the parallel deployment of different Cry toxins (Shelton et al., 2000). Mixtures of Cry toxins can be attained: by inter-planting two cultivars, each with a different genetic basis for resistance (mosaic planting); by sequentially planting two cultivars; by gene pyramiding, where at least two Cry toxins are expressed in the same cultivar (Cao et al., 2002; Zhao et al., 2003), and by mixing with non-Cry insecticidal genes such as gna (snowdrop lectin) (Maqbool et al., 2001) and CpT1 encoding a cowpea trypsin inhibitor (Zhao et al., 1999).

A colony of H. virescens selected with Cry1Ac pro-toxin attained over 10,000-fold resistance to Cry1Ac but was not resistant to Cry1Ca and Cry1Ba (Gould et al., 1995). The genes coding for Cry1Ac and Cry1Ca were therefore cloned here, separately and together, for expression in Escherichia coli, and toxicities of the clones were tested against three species of agricultural pests.

139

140

Materials and Methods Bacterial strains, plasmids and genes Strain XL-Blue MRF´ of E. coli (Stratagene, La Jolla, CA) was used throughout as a host. Previously constructed plasmid pBS SK/N, a derivative of pBluescript II SK+ (Wu et al., 1997) harboring the early T7 promoter PA1 that utilizes the usual E. coli RNA polymerase (Deuschle et al., 1986), served for cloning cry1Ac and cry1Ca and their combination (Table 1). cry1Ac was isolated from Bt subsp. kurstaki-4D4 (original code HD-73) and cry1Ca from Bt subsp. aizawai-4J4 (original code HD11).

Table 1. Plasmids expressing combinations of δ-endotoxin genes from B. thuringiensis.

Plasmid Description (genes cloned from Bt)

pBt-1A derivative of pBS-SK/N (cry1Ac)

pBt-1C derivative of pBS-SK/Na (cry1Ca)

pBt-1CA derivative of pBS-SK/N (cry1Ca and cry1Ac)

a, pBS-SK/N, derivative of pBluescript II SK+ with 716 bp XbaI-XhoI fragment carrying PA1 promoter from pUHE-24S (Wu et al., 1997).

Polymerase chain reaction Amplification was carried out with the high fidelity Vent DNA polymerase (New England Biolabs) in a Biometra TGradient thermocycler (Biometra GmbH, Göttingen, Germany) set to the following 30 reaction cycles: 40 s at 94°C, 40 s at 55°C and 4 min at 72°C.

Three primer pairs (Table 2) were employed to obtain cry1Ac (31-mer Cry1A-NcoI-d and 34-mer Cry1A-SalI-r or 36-mer Cry1A-SalI-d and 28-mer Cry1A -SalII-r) and cry1Ca (28-mer Cry1C-NcoI-d and 30-mer Cry1C-SalI-r).

Table 2. List of primers designed and used in this investigation.

Description/Namea Sequence (from 5' to 3')b

Cry1A-NcoI-d GAGATGGAGGTAACCCATGGATAACAATCCG

Cry1A-SalI-r GTAATGCTGCTCGTCGACAATTGATTTGAAAACG

Cry1A-SalI-d GAGTCATATGTTTTAAATTGTCGACATGAAAAACAG

Cry1A-SacII-r GTAATGCTGCCCGCGGACAATTGATTTG

Cry1C-NcoI-d CGGAGGTATTCCATGGAGGAAAATAATC

Cry1C-SalI-r CTTTATTTGTCGACGTTACATTTTATAACG

a d - direct primer; r - reverse primer. b The restriction sites are bold-faced.

141

Cloning of cry1Ac and cry1Ca, alone or in combination The amplicons containing cry1Ac and cry1Ca were separately digested by NcoI/SalI and inserted, each into pBS SK/N (Wu et al., 1997) under the PA1 promoter, to yield clones pBt-1A and pBt-1C, respectively. The SalI/SacII-digested fragment with cry1Ac was inserted into pBt-1C downstream from cry1Ca to obtain pBt-1CA with both genes under PA1 promoter (Table 1). Insects and bioassays for larvicidal activities A colony of P. gossypiella and one of H. armigera were reared on an artificial diet (Manduca-Heliothis Premix, Stonefly, TX, USA) under standard controlled conditions: 27±2°C, 60% RH and photoperiod of 14:10 h (L:D). A colony of S. littoralis was maintained on castor bean leaves at 25±2°C, 60% RH with photoperiod of 14:10 h (L:D).

Cells of the recombinant E. coli were harvested by centrifugation after 4 hrs of induction and re-suspended in distilled water. For P. gossypiella, samples were thoroughly blended into the diet to achieve the required concentrations (Tabashnik et al., 2000). For each treatment, ten plastic cups (50 ml) with 10 g of diet and 10 Lepidoptera eggs in each were incubated at 27±2°C, 60% RH and photoperiod 14:10 h (L:D), and the number of surviving larvae and their weights were recorded after 15 days.

For H. armigera and S. littoralis, a disk (5-cm diameter) of fresh cotton leaf was immersed in a test solution for 10 s, allowed to dry at ambient temperature for 2 h, and placed in a Petri dish containing moistened filter paper. Two 1st-instar larvae were placed in each

dish, and each treatment was repeated ten times. The dishes were incubated at 27±2°C, 60% RH and photoperiod of 14:10 h (L:D). Mortality, larvae weights and leaf damage levels were determined after 5 days.

Values of LC50 and LC90 (concentrations of cells that kill 50% and 90% of the exposed populations, respectively) were determined by using EPA Probit analysis programme with at least six doses. All bioassays were performed at least thrice for each concentration.

Values of Synergy Factor (SF) were calculated according to Tabashnik (1992), assuming that all proteins were produced equally in the recombinants, by dividing the theoretical LC50 or LC90 (estimated from the cell number harmonic mean of the LC50 or the LC90 of the clones producing the component proteins) by the respective observed values. Results and Discussion Interactions between Cry1Ac and Cry1Ca against lepidopteran species The genes cry1Ca and cry1Ac were cloned separately and together to derive pBt-1C, pBt-1A and pBt-1CA, respectively (Tables 1 and 2). As anticipated, pBt-1C displayed some toxicity against larvae of S. littoralis (LC50 of 9.8×108 cells ml-1; Table 3) but none against P. gossypiella and H. armigera. In contrast, pBt-1A exhibited the highest level of toxicity against P. gossypiella (LC50 of 0.27×108 cells g-1) a relatively high level against H. armigera larvae (LC50 of 1.6×108 cells ml-1), but none against S. littoralis (Table 3). This was also consistent with the amount of expressed proteins (not shown). However, pBt-1CA that produced both toxins was not toxic to H. armigera and showed a similar level of toxicity (LC50 of 0.4×108 cells g-1) as that of pBt-1A against P. gossypiella larvae (Table 3). The low toxicity of this clone may be explained by the interaction between Cry1Ca, which is not toxic to P. gossypiella and H. armigera larvae, and Cry1Ac, which is toxic against those pests (Shao et al., 2001). Consistently, some combinations of co-expressed cry genes can affect the solubility and composition of inclusion bodies and hence of toxicity properties (Aronson,

142

1995). The downstream position of cry1Ac in the construct may reduce its expression level and subsequently distort the appropriate stoichiometry. This sequence of the two genes, on the other hand, exhibited a rather high level of toxicity against S. littoralis larvae (LC50 of 0.12×108 cells ml-1) compared to pBt-1C (LC50 of 9.8×108 cells ml-1) and to pBt-1A, which was not toxic, and an SF of 164 (Table 3). Highest toxicities were observed against larvae of both, S. exigua and H. armigera, in the combination of Cry1Aa and Cry1Ca (produced in acrystalliferous Bt 4Q7) at a ratio of 1:1 (Xue et al., 2005). The differences in toxicity levels of pBt-1CA against larvae of H. armigera, P. gossypiella and S. littoralis may result from the different ensemble of receptors in these pests, and various modes of interactions of the toxins in the different environments that exist in their guts. Table 3. Larvicidal activities of the transgenic E. coli clones against 3 lepidopteran speciesa.

Larvicidal activity againstb

P. gossypiellac H. armigerac S. littoralisc

E. coli strains

Bt genes

LC50 LC90 LC50 LC90 LC50 LC90

pBt-1A cry1Ac 0.27 (0.04-5.3)

1.47 (0.2-8.2)

1.57 (0.6-3.9)

37.05 (11-56)

ND ND

pBt-1C cry1Ca ND ND ND ND 9.81 (4.1-17.1)

Not reached

pBt-1CA cry1Ca, cry1Ac 0.4 (0.14-0.76)

4.8 (2-28)

ND ND 0.12 (0.02-0.3)

Not reached

a, The genes are expressed under the early T7 promoter (PA1). b, Values of LC50 and LC90 are averages of four bioassays. Numbers in parentheses are 95%

confidence limits, as determined by probit analysis. ND, no toxicity detected, even at a concentration of 1×1010 cells ml-1 or g-1.

c , Larvicidal activities are expressed in terms of 108 cells per either ml (for H. armigera and S. littoralis) or gram of the diet (for P. gossypiella).

Brush Border Membrane Vesicles (BBMV) of S. littoralis larvae were permeabilize to KCl by Cry1Ca but not by Cry1Ac, which is only marginally active (Escriche et al., 1998). Swelling experiments suggest that the very low levels of activity of Cry1Ac in S. littoralis may be due to inability to carry perforation even if binding can take place. This is consistent with the observation that Cry1Ab and Cry1Ac compete for the same binding site in S. frugiperda larvae, whereas Cry1Ba and Cry1Ca compete for another (Rang et al., 2004). It is noteworthy that the maximum osmotic swelling rate was much higher when the BBMV of Manduca sexta larvae were incubated with both Cry1Ac and Cry1Ca, an indication that the number of pores formed by each of the toxins is limited by the availability of their respective receptors (Mélanie Fortier et al., 2007).

143

Acknowledgements Thanks are due to Dr. Rami Horowitz and his team at the Agricultural Research Organization (ARO), Professor Sammy Boussiba and his at the Institute for Desert Research, and Mrs. Monica Einav at BGU’s Life Sciences Department, for their fruitful collaboration and help in various stages of the research. This investigation was partially supported by grants of the Chief Scientist of the Israel Ministry of Agriculture (No. 857-0481-03) and from the United States-Israel Binational Science Foundation (B.S.F.), Jerusalem, Israel (No. 2001-042). References Aronson, A.I. 1995: The protoxin composition of Bacillus thuringiensis insecticidal

inclusions affects solubility and toxicity. Appl. Environ. Microbiol. 61: 4057-4060. Cao, J., Zhao, J.-Z., Tang, J.D., Shelton, A.M. & Earle, E.D. 2002: Broccoli plants with

pyramided cry1Ac and cry1C Bt genes control diamondback moths resistant to Cry1A and Cry1C proteins. Theor. Appl. Genet. 105: 258-264.

Deuschle, U., Kammerer, W., Gentz, R. & Bujard, H. 1986: Promoters of Escherichia coli: a hierarchy of in vivo strength indicates alternate structures. EMBO J. 5: 2987-2994.

Escriche, B., de Decker, N., van Rie, J., Jansens, S. & van Kerkhove, E. 1998: Changes in permeability of brush border membrane vesicles from Spodoptera littoralis midgut induced by insecticidal crystal proteins from Bacillus thuringiensis. Appl. Environ. Microbiol. 64: 1563-1565.

Fortier, M., Vachon, V., Marceau, L., Schwartz, J.-L. & Laprade, R. 2007: Kinetics of pore formation by the Bacillus thuringiensis toxin Cry1Ac. Biochim. Biophys. Acta 1768: 1291-1298.

Gould, F., Anderson, A., Reybolds, A., Baumgarber, L. & Moar, W.J. 1995: Selection and genetic analysis of a Heliothis virescens (Lepidoptera: Noctuidae) strain with high levels of resistance to some Bacillus thuringiensis toxins. J. Econ. Entomol. 88: 1545-1559.

Maqbool, S.B., Riazuddin, S., Loc, N.T., Gatehouse, A.M.R., Gatehouse, J.A. & Christou, P. 2001: Expression of multiple insecticidal genes confers broad resistance against a range of different rice pests. Mol. Breed. 7: 85-93.

Moar, W.J., Pusztai-Carey, M., Van Faassen, H., Bosch, D., Frutos, R., Rang, C., Luo, K. & Adang, M.J. 1995: Development of Bacillus thuringiensis Cry1C resistance by Spodoptera exigua (Hübner) (Lepidoptera: Noctuidae). Appl. Environ. Microbiol. 61: 2086-2092.

Rang, C., Bergvingson, D., Bohorova, N., Hoisington, D. & Frutos, R. 2004. Competition of bacillus thuringiensis Cry1 toxins for midgut binding sites: a basis for the development and management of transgenic tropical maize resistant to several stemborers. Curr. Microbiol. 49: 22-27.

Sayyed, A., Haward, R., Herrero, S., Ferré, J. & Wright, D.J. 2000: Genetic and biochemical approach for characterization of resistance to Bacillus thuringiensis toxin Cry1Ac in a field population of the diamondback moth, Plutella xylostella. Appl. Environ. Microbiol. 66: 1509-1516.

Schnepf, E., Crickmore, N., Van Rie, J., Lereclus, D., Baum, J., Feitelson, J., Zeigler, D.R. & Dean, D.H. 1998: Bacillus thuringiensis and its pesticidal crystal proteins. Microbiol. Mol. Biol. Rev. 62: 775-806.

144

Shao, Z., Liu, Z. & Yu, Z. 2001: Effects of the 20-Kilodalton helper protein on Cry1Ac production and spore formation in Bacillus thuringiensis. Appl. Environ. Microbiol. 67: 5362-5369.

Shelton, A.M., Tang, J.D., Roush, R.T., Metz, T.D. & Earle, E.D. 2000: Field tests on managing resistance to Bt-engineered plants. Nat. Biotechnol. 18: 339-342.

Tabashnik, B.E. 1992: Evaluation of synergism among Bacillus thuringiensis toxins. Appl. Environ. Microbiol. 58: 3343-3346.

Tabashnik, B.E. 1994: Evaluation of resistance to Bacillus thuringiensis. Ann. Rev. Entomol. 39: 47-49.

Tabashnik, B.E., Finson, N., Groeters, F.R., Moar, W.J., Johnson. M.W., Luo, K. & Adang, M.J. 1994: Reversal of resistance to Bacillus thuringiensis in Plutella xylostella. Proc. Natl. Acad. Sci. USA 91: 4120-4124.

Tabashnik, B.E., Liu, Y.-B., de Maagd, R.A. & Dennehy, T.J. 2000: Cross-resistance of pink bollworm (Pectinophora gossypiella) to Bacillus thuringiensis toxins. Appl. Environ. Microbiol. 66: 4582-4584.

Wu, X., Vennison, S.J., Liu, H., Ben-Dov, E., Zaritsky, A. & Boussiba, S. 1997: Mosquito larvicidal activity of transgenic Anabaena strain PCC 7120 expressing combinations of genes from Bacillus thuringiensis subsp. israelensis. Appl. Environ. Microbiol. 63: 1533-1537.

Xue, J.-L., Cai, Q.-X., Zheng, D.-S. & Yuan, Z.-M. 2005: The synergistic activity between Cry1Aa and Cry1C from Bacillus thuringiensis against Spodoptera exigua and Helicoverpa armigera. Lett. Appl. Microbiol. 40: 460-465.

Zhao, J.-Z., Fan, Y.L., Fan, X.L., Shi, X.P. & Lu, M.G. 1999: Evaluation of transgenic tobacco expressing two insecticidal genes to delay resistance development of Helicoverpa armigera. Chinese Sci. Bull. (English edition) 44: 871-1874.

Zhao, J.-Z., Cao, J., Li, Y., Collins, H.L., Roush, R.T., Earle, E.D. & Shelton, A.M. 2003: Transgenic plants expressing two Bacillus thuringiensis toxins delay insect resistance evolution. Nat. Biotechnol. 21: 1493-1497.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 145-148

Assessing the effects of Bt-maize pollen on Typhlodromus pyri (Acari: Phytoseiidae) Rostislav Zemek, Zuzana Vávrová Institute of Entomology, Biology Centre AS CR, Branisovska 31, CZ-370 05 Ceske Budejovice, Czech Republic (E-mail: rosta@acarus.entu.cas.cz); University of South Bohemia, Branisovska 31, CZ-370 05 Ceske Budejovice, Czech Republic (E-mail: zuzavavrova@seznam.cz) Abstract: Transgenic maize carrying the lepidopteran-active cry1Ab gene from Bacillus thuringiensis kurstaki proved to be effective in the control of stem borers. Since the Cry toxin is also expressed in maize pollen it could potentially be harmful for pollen-feeding non-target arthropods. We assessed the impact of MON 810 maize on the performance of Typhlodromus pyri Scheuten, an omnivorous predatory mite which is considered important for biological control of many phytophagous mites. Its ability to utilize a wide range of food including pollen of various plant and tree species has an important role for its success at low prey density. Various parameters including longevity, developmental time and fecundity were measured in laboratory experiments when predatory mites were offered Bt or non-Bt maize pollen as a food source. The obtained results revealed that predatory mites survived, developed and reproduced well on Bt maize pollen and no significant differences compared to mites reared on non-Bt maize pollen were found. We can thus conclude that pollen from Bt maize has no detrimental effect on T. pyri. Key words: risk assessment, Cry1Ab, predatory mites Introduction During its sporulation the gram-positive soil bacterium Bacillus thuringiensis (Bt) produces a parasporal protein inclusion crystal composed by one or more delta-endotoxins with specific insecticidal activity. The incorporation of B. thuringiensis delta-endotoxin cry genes into the plant genome and its sufficient expression in plant tissues have been used with the objective of protecting the plant from insect attack by targeting the digestive system of insect pests (Schuler et al., 1998). These so-called Bt plants are commercialy available since the mid 1990s and the percentage of the global agricultural area where they are grown is increasing considerably (James, 2006). Since most commercial Bt maize events express the Cry protein in the pollen which can be deposited also on plants in surrounding fields, pollen feeding non-target arthropods may be exposed to the insecticidal compounds.

Plant-inhabiting predaceous mites of the family Phytoseiidae are prominent in biological/integrated pest management of phytophagous mites and insect. Several species of the family are omnivorous predators which also feed on non-prey food including pollen. One of them, Typhlodromus pyri Scheuten, is widely used for biological control in orchads and vineyards. Its ability to utilize pollen of various plants and trees as an alternative food is known to be very important for its success at low prey densities (Zemek, 2005). The study reported here focused on the effects of pollen of the transgenic maize event MON810 expressing Cry1Ab protein targeted agains stem borers (Koziel et al., 1993) on the performance of T. pyri.

145

146

Material and Methods Transgenic (MON 810) maize plants expressing the lepidopteran-active Cry1Ab toxin and their corresponding non-transformed near isoline were obtained from Monsanto company. The plants were grown in a greenhouse. The pollen was collected using paper bags and stored at -18 °C.

Overwintered T. pyri females used in the experiments were collected in apple orchards in Chelčice, Czech Republic. The females were individually kept in arenas made of 2x2 cm HD-PE black plastic (thickness 0.023 mm) laid on water-saturated plastic foam in a Petri dish. An approx. 2 mm wide sticky barrier (Biocont Laboratory, Brno) was made close to the arena edges to prevent mites from escaping. Water was provided by a wet cotton wick. A roof-shape plastic served as a shelter and an oviposition site for females. The arenas were placed in a climatized cabinet at 20±1°C and 18L:6D photoperiod. Humidity was not controlled but it was expected to be at least 70% in arenas surrounded by water in Petri dishes. The rearing units were inspected daily using a dissection microscope (magnification 16x and 25x) and number of females alive and number of eggs laid were recorded. Once females started to reproduce, their progeny was transferred into new arenas where juveniles were fed by the same pollen until their adulthood. New pollen was added daily and was always ad libitum. Mites were transferred to new rearing units weekly to keep the culture clean and free of fungi. Two parallel series of experiments, each starting with 24 females, were conducted: one with Bt pollen and one with non-Bt pollen. Presence of the Bt protein in pollen used was confirmed at the end of experiments with the QuickStix kit from Envirologix.

The obtained data were analysed using survival analysis (Cox, 1972) and Student’s t-test. Results and Discussion The predatory females fed on Bt and non-Bt pollen survived in average (±SD) 48.8±25.1 and 55.0±19.7 days, respectively. Survival analysis revealed no significant differences (P=1.3). None of the reproduction parameters of T. pyri fed on Bt maize pollen was significantly different from those of T. pyri fed on non-Bt maize pollen (Table 1). Similarly, no significant differences in developmental time between progeny fed on Bt and non-Bt pollen were found (Table 2). Table 1. The effect of Bt maize pollen on reproduction parameters (means±SD) of T. pyri. Numbers in parentheses indicate number of individuals. Fecundity was calculated only for females which completed their oviposition, i.e. excluding those which were accidentally lost.

Parameter Bt non-Bt P (t-test) Preoviposition period (days) 21.6±9.0 (21) 18.9±6.6 (23) 0.276 Fecundity (eggs per female) 7.9±4.0 (17) 9.5±3.6 (19) 0.628 Oviposition rate (eggs per day) 0.18±0.05 (21) 0.19±0.05 (23) 0.627

147

Table 2. The effect of Bt maize pollen on developmental time (means±SD in days) of T. pyri juvenile stages. Numbers in parentheses indicate number of individuals.

Developmental stage Bt non-Bt P (t-test)

Egg 4.9±1.0 (152) 4.7±0.9 (179) 0.068 Larva 1.2±0.2 (151) 1.3±0.2 (174) 0.330 Protonymph 5.0±0.6 (108) 4.8±0.7 (132) 0.320 Deutonymph 5.0±0.5 (91) 5.0±0.6 (111) 0.350

Several authors stressed the importance of alternative food and namely pollen for T. pyri (Collyer, 1964; Solomon, 1982; Engel, 1990). Access to pollen enables this predatory mite not only to survive, but also to reproduce during periods when the density of its principal prey is low.

The present study addressed the question whether Bt-maize pollen has any effect on the performance of T. pyri in laboratory experiments. To our knowledge, the effect of Bt-transgenic plants on predatory mites has only been studied using transgenic eggplants, Solanum melongena L., expressing the coleopteran-active Cry3Bb toxin (Zemková Rovenská et al., 2005) and using transgenic maize (Bt 11) expressing lepidopteran-active Cry1Ab toxin (Obrist et al., 2006). The results from the first study indicate that spider mites fed on Bt eggplant are less preferred by their predator Phytoseiulus persimilis A.-H. when compared to mites fed with non-transformed eggplants (choice experiments). Whether the observed effect was due to the Cry3Bb protein, or due to other changes in primary or secondary metabolites in the transgenic eggplants is unclear and remains to be elucidated. In contrast, Obrist et al. (2006) found that there were no statistically significant prey-mediated effects of Bt-maize on Neoseiulus cucumeris (Oudemans). However, when the predatory mites were fed with pollen, some life-table parameters differed significantly between Bt and non-Bt maize pollen. The authors did not attribute this difference to the Bt toxin because its content in pollen was relatively low compared to the content of toxin in spider mites.

The present paper failed to find any statistically significant differences between T. pyri fed on Bt and non-Bt maize pollen and we can thus conclude that the obtained results indicate no detrimental effect of Bt maize pollen on this phytoseiid mite species. Acknowledgements This work was supported by the Institute of Entomology project and Grant Agency of the Academy of Sciences of the Czech Republic (A6007303). The seeds of MON 810 maize were kindly provided by Monsanto company. The authors thank Dr. Jörg Romeis for his comments on an earlier version of the manuscript. Mrs. Jana Jabůrková is thanked for her technical assistance. References Collyer, E. 1964: The effect of an alternative food supply on the relationship between two

Typhlodromus species and Panonychus ulmi (Koch) (Acarina). Entomol. Exp. Appl. 7: 120-124.

Cox, D.R. 1972: Regression models and life tables. J. R. Statist. Soc. B 34: 187-203. Engel, R. 1990: Alternative prey and other food resources of the phytoseiid mite Typhlodromus

pyri (Scheuten). IOBC/WPRS Bulletin 13(7): 124-127.

148

James, C. 2006: Global Status of Commercialized Biotech/GM crops: 2006. ISAAA Brief No. 35. International Service for the Acquisition of Agri-Biotech Applications, Ithaca, New York.

Koziel, M.G., Beland, G.L., Bowman, C., Carozzi, N.B., Crenshav, R., Crossland, L., Dawson, J., Desai, N., Hill, M. Kadwell, S., Launis, K., Lewis, K., Maddox, D., Mc Pherson, K., Meghji, M.R., Merlin, E., Rhodes, R., Warren, G., Wright, M. & Evola, S.V. 1993: Field performance of elite transgenic maize plants expressing an insecticidal protein derived from Bacillus thuringiensis. Biotechnol. 11: 194-200.

Obrist, L.B., Klein, H., Dutton, A. & Bigler, F. 2006: Assessing the effects of Bt maize on the predatory mite Neoseiulus cucumeris. Exp. Appl. Acarol. 38: 125-139.

Schuler, T.H., Poppy, G.M., Kerry, B.R. & Denholm, I. 1998: Insect resistant transgenic plants. Trends Biotechnol. 16: 168-175.

Solomon, M.G. 1982: Phytophagous mites and their predators in apple orchards. Ann. Appl. Biol. 101: 201-203.

Zemek, R. 2005: The effect of powdery mildew on the number of prey consumed by Typhlodromus pyri (Acari: Phytoseiidae). J. Appl. Entomol. 129: 211-216.

Zemková Rovenská, G., Zemek, R., Hilbeck, A. & Schmidt, J.E.U. 2005: Altered host plant preference of Tetranychus urticae and prey preference of its predator Phytoseiulus persimilis (Acari: Tetranychidae, Phytoseiidae) on transgenic Cry3Bb-eggplants. Biol. Control 33: 293-300.

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 149-155 Non-target arthropod risk assessment of insect-resistant GM crops Jörg Romeis Agroscope Reckenholz-Tänikon Research Station ART, Reckenholzstr. 191, 8046 Zurich, Switzerland (E-mail: joerg.romeis@art.admin.ch) Abstract: Transgenic insecticidal crops have the potential to pose risks to non-target organisms. These risks need to be addressed as part of the environmental risk assessment that precedes the commercialization of any novel transgenic crop. An international initiative has been launched to develop a scientifically-sound, generic, and pragmatic approach to assess the risks to terrestrial non-target arthropods. The basis for this work is the widely-established and effective tiered testing approach from regulatory toxicology. The basic principles of this approach are described. These may provide guidance to countries that are currently developing their own non-target risk assessment guidelines and help to harmonize regulatory requirements in different regions. Key words: biosafety, non-target arthropods, risk assessment, tiered approach, transgenic crops Background Insect-resistant genetically modified (IRGM) crops that express Cry proteins derived from the soil bacterium Bacillus thuringiensis (Bt) have been grown on a steadily increasing area worldwide since their first introduction in 1996, reaching more than 32 million hectares in 2006 (James, 2006). A number of crops expressing novel insecticidal proteins are also under development and expected to reach the market stage in the near future. Like conventional agricultural pest control products, one of the risks associated with the growing of IRGM crops is their potential impact to non-target organisms, including a range of arthropod species that fulfil important ecological functions such as biological control, pollination and decomposition. Potential non-target risks are thus assessed as part of the environmental risk assessment (ERA), prior to the cultivation of any transgenic crop.

Regulations and guidelines exist that provide general guidance on conducting an ERA of transgenic plants. However, there is still a need for detailed descriptions of non-target risk assessment procedures, for development of rigorous criteria for the development of non-target species that need to be tested, and for establishment of test methods that can apply to different regions. This need has been identified, among others, at a workshop held during the first full meeting of the IOBC/WPRS working group “GMOs in Integrated Plant Production” in 2003 in Prague (Romeis, 2004). As a follow-up, a special workshop was organized adjacent to the second full working group meeting in 2005 in Lleida, Spain. The workshop was attended by participants from public research institutes, private industry and regulatory authorities. The participants were invited based on their long standing experience in (GMO) risk assessment and regulation. The aim of the workshop was to identify areas where members of the IOBC/WPRS WG could facilitate the development of guidelines related to non-target risk assessment of IRGM crops. It appeared that there is a clear need to focus the activities on the following three areas: (i) Development of a generic risk assessment process for non-target organisms (ii) Definition of criteria for the selection of non-target organisms to be assessed (iii) Development of standard test methods for selected non-target species

149

150

A workshop protocol with a list of participants has been included in the proceedings from the Lleida meeting (Romeis, 2006).

Our group and aims In December 2005 the initiative on “Non-target risk-assessment and regulation” was formally accepted by the IOBC/WPRS Council and a first workshop was organized in Engelberg, Switzerland, in May 2006 with the aim of establishing generic and scientifically rigorous ERA guidelines for IRGM crops, focusing on terrestrial non-target arthropods (NTA’s). This should help regulators in their task to define the problem to be addressed in the ERA as effectively as possible and to formulate this into a set of questions that are amenable to scientific inquiry, supporting regulatory decision-making within a reasonable time frame.

The expert group consists of European and North American scientists from public research institutes, the agricultural biotechnology industry, regulatory agencies, and a commercial testing laboratory (see list of participants below). The group thus has experience in the application of tiered risk assessment from a research and regulatory perspective. We have identified the most valuable elements from within a variety of guidance documents and procedures and we have distilled the lessons that have accumulated as a result of the institutional experience of the working group members. Presentations and Publications We were able to first present our approach at the 9th International Symposium on the Biosafety of Genetically Modified Organisms (ISBGMO) organized by the International Society for Biosfety Research (ISBR) in September 2006 on Jeju Island, South Korea (Romeis et al., 2006). Regulators could directly be addressed at a scientific colloquium organized by the European Food Safety Authority (EFSA) in Tabiano, Italy, in June 2007 (EFSA, 2007). Other presentations include those at the V Brazilian Biosafety Congress in Ouro Preto in September 2007 (Romeis et al., 2007) and at the Annual Meeting of the Entomological Society of America in December 2007 within a symposium entitled “Harmonizing laboratory methods to evaluate potential effects of genetically-engineered crops on non-target organisms”. Finally the proposed approach was published in Nature Biotechnology in February 2008 (Romeis et al., 2008). Our Non-Target Risk Assessment Approach The developed consensus approach consists of an adaptation of the tiered approach to risk assessment that is used internationally within regulatory toxicology and environmental sciences (Touart & Maciorowski, 1997), and versions of which are already in use in established regulatory systems for GM crops (e.g., EC 2001, 2002; CFIA, 2004; EFSA, 2006; Rose, 2007). The approach has a strong focus on the formulation and testing of clearly stated risk hypotheses, making maximum use of available data and using formal decision guidelines to progress between testing stages (or tiers). Problem formulation During the problem formulation stage, meaningful differences between the IRGM plant and its non-GM counterparts are identified in order to focus the ERA on the areas of greatest concern or uncertainty (Raybould, 2006). This includes establishing the level of ‘familiarity’ (that is the similarities in ecologically relevant characteristics) between the IRGM crop and non-transformed crop. The problem formulation furthermore considers the specifics of the

151

mode of action of the expressed insecticidal protein, the spectrum of activity and susceptibility, mode of expression, and relevant spatial and temporal exposure profiles. Additionally, it must also take into account ecological considerations that might affect the nature and extent of possible environmental impacts. In all cases, descriptions of plant characteristics (e.g., macro- and micro-nutrient composition, content of important toxicants and anti-nutrients, and morphological and agronomic plant characteristics) are made with reference to plants that are generally regarded as environmentally ‘acceptable’ to identify meaningful differences that may need to be addressed in the risk assessment. This assessment (generally referred to as the concepts of ‘familiarity’ and ‘substantial equivalence’) serve as a starting point to focus the ERA process on potential stressors of concern (OECD, 1993; Kuiper et al., 2001). On this basis, the problem formulation then identifies assessment endpoints reflecting management goals and the scale and nature of the receiving ecosystem that is being considered. If a lack of significant differences between the IRGM plant and its comparators is established, the ERA can emphasize the effects of the insecticidal protein. A typical risk hypothesis resulting from this procedure may be that the insecticidal protein does not cause any harm to NTA’s at the concentration expressed in the field.

Regardless of where in the world the ERA is conducted, the problem formulation approach should be very similar, using similar information that is modified by local cropping system information. The framework A tiered risk assessment is recognized as being the most appropriate and rigorous approach to assess non-target affects from both scientific and regulatory standpoints (e.g., Garcia-Alonso et al., 2006). Both hazard and exposure can be evaluated within different levels or “tiers” that progress from worst-case hazard and exposure to more realistic scenarios. Lower tier tests serve to identify potential hazards, and they are generally conducted in the laboratory to provide high levels of replication and study control, which increase the statistical power to test risk hypotheses. Where potential hazards are detected in these early tier tests, additional information is required. In these cases, higher tier tests can serve to confirm whether an effect might still be detected at more realistic rates and routes of exposure. Higher tier studies including semi-field or field-based tests offer greater environmental realism, but they may have lower statistical power. These tests are only triggered when early tier studies in the laboratory indicate potential hazards at environmentally relevant levels of exposure. In exceptional cases, higher tier studies may be conducted at the initial stage when early tier tests are not possible, for example plant tissue might be used because purified toxin is not available. Higher levels of replication or repetition may be needed to enhance statistical power in these circumstances.

In cases where a potential hazard is detected in a lower tier test, the tiered approach provides the flexibility to undertake further lower tier tests in the laboratory to increase the taxonomic breadth or local relevance of test species. Depending on the nature of the effect, one may also progress to higher tier testing, particularly in cases where there is no previous experience with the crop or toxin under investigation.

Movement between tiers is based on the sufficiency of information that is available. If sufficient data and experience from toxicological testing and exposure analyses are available to characterize the potential risk as being acceptable, then there is no need to undertake additional testing. The process is designed to optimize the use of resources and to identify and define sources of potential risk with high scientific rigour.

152

Species selection For practical reasons, only a small fraction of all possible terrestrial arthropods can be considered for regulatory testing. It is therefore necessary to select appropriate species to serve as surrogates for ecologically and economically important NTA’s that can be tested under worst-case conditions in the laboratory. Species should be chosen to represent different ecological functions such as herbivory, pollination of cultivated and wild plants, predation and parasitism of pest organisms, and decomposition in the soil. In order to reflect biogeographical variation, it is crucial to determine what taxa are likely to occur in the cropping systems where the transgenic plant is expected to be grown. Another important source of information that serves as a basis for selecting relevant surrogate species is the information on the insecticidal protein (specificity, mode of expression and exposure profile) that is accumulated during problem formulation. The information collected in these previous steps will direct the selection of representative NTA’s from a proposed set of species that capture key ecological functions. Criteria such as amenability to testing, availability of test methods and unambiguous taxonomic recognition are crucial for non-target testing. Based on these criteria, a list of NTA species that are representatives of those living in the crop and in adjacent non-crop habitats is proposed. As a result of this process, test protocols for species that are of high relevance in particular regions may need to be developed. The application of the surrogate species concept enhances the transferability of data from lower tier tests to a wide range of regions and crops. Study design Hazard assessment tests are usually conducted using elevated protein doses in the laboratory, following standardized testing protocols. This assures a high level of confidence in the conclusions drawn from the data and applicability for further ERA’s. Prior to testing, the objectives of the individual studies need to be defined, and specific measurement endpoints described that address the risk hypotheses (and are related to assessment endpoints). Appropriate measurement endpoints include life-cycle parameters such as mortality or fecundity, which can easily be evaluated and related to assessment endpoints such as population size of a certain NTA. Other endpoints such as body mass or behavioral effects may be used; however, risk assessors should agree how to interpret these data if they have a less direct connection with population statistics

Testing protein concentrations that are several times higher than those that will be seen in the field increases the likelihood that a hazard will be detected should one be present. All tests should adopt quality control criteria that help to validate the test system. For example, for lower tier tests these may include: (i) a requirement for low negative control mortality, (ii) the use of a positive toxic control to confirm that the test system is working effectively, (iii) homogeneity of the test material to ensure uniformity of exposure, (iv) stability of the insecticidal compound throughout the bioassay period, and (v) sufficient statistical power for the testing of risk hypotheses.

Higher tier tests that are, for example, conducted in the field are more realistic but highly complex. They have a high intrinsic uncertainty for showing hazards but more certainty for showing whether hazards pose a risk. Higher tier studies should thus only be conducted when they can further reduce uncertainty in the risk assessment, and only when justified by detection of potentially adverse effects in the lower tiers of testing. Thus, effective tiered processes prevent costly and unnecessary testing.

153

Conclusions The tiered approach presented here ensures rigorous testing of clearly stated and relevant risk hypotheses and thereby minimizes collection of data that are irrelevant to the risk assessment. It thus provides a rigorous and effective basis for determining the potential of IRGM plants to adversely affect NTA’s, from both scientific and regulatory standpoints.

We believe that our consensus approach can provide guidance to regulatory authorities that are currently developing their own NTA risk assessment guidelines for IRGM crops and to help harmonize regulatory requirements among different countries and different regions of the world. Outlook We have identified three main areas for further activities: i) Definition of threshold values that trigger the advance to higher tiers in NTA risk

assessment. ii) A list of surrogate species needs to be compiled to serve as a basis from which to collect

the most appropriate species for testing. iii) More standardized, validated test protocols for surrogate test species need to be

developed. Participants The following colleagues have participated in this activity since it’s establishment in December 2005. Detlef Bartsch, Federal Office of Consumer Protection and Food Safety (BVL), Mauerstrasse

39-42, 10117 Berlin, Germany Franz Bigler, Agroscope Reckenholz-Tänikon Research Station ART, Reckenholzstr. 191,

8046 Zurich, Switzerland Marco P. Candolfi, RCC Laboratories India Private Ltd., Genome Valley, Turkapally,

Shameerpet, Ranga Reddy District, Hyderabad 500078, India (present address: BASF SE, APD/EE, LI 425, 67117 Limburgerhof, Germany)

Marco M.C. Gielkens, National Institute for Public Health and the Environment, Expertise Centre for Substances, PO Box 1, 3720 BA Bilthoven, Netherlands

Susan E. Hartley, Department of Biology and Environmental Science, University of Sussex, Falmer, Brighton BN1 9QG, UK

Richard L. Hellmich, USDA–ARS, Corn Insects and Crop Genetics Research Unit and Department of Entomology, Iowa State University, 110 Genetics Laboratory c/o Insectary, Ames, IA 50011, USA

Joseph E. Huesing, Regulatory Sciences, Monsanto Company, 800 N. Lindbergh Blvd., St. Louis, MO 63167, USA

Paul C. Jepson, Integrated Plant Protection Center and Department of Environmental and Molecular Toxicology, Oregon State University, Cordley Hall, 2040, Corvallis, OR 97331-2907, USA

Raymond Layton, Regulatory Sciences, DuPont Crop Genetics, 7250 NW 62nd Ave., PO Box 552, Johnston, IA 50131-0052, USA

Hector Quemada, Program for Biosafety Systems, Department of Biology, Calvin College, 1726 Knollcrest Circle, S.E., Grand Rapids, MI 49546-4403, USA

154

Alan Raybould, Syngenta, Jealott’s Hill International Research Centre, Bracknell, Berkshire RG42 6EY, UK

Jörg Romeis, Agroscope Reckenholz-Tänikon Research Station ART, Reckenholzstr. 191, 8046 Zurich, Switzerland

Robyn I. Rose, USDA–APHIS Biotechnology Regulatory Services, 4700 River Rd, Unit 146, Riverdale, MD 20737, USA

Joachim Schiemann, Julius Kuehn Institute, Federal Research Centre for Cultivated Plants (JKI), Institute for Biosafety of Genetically Modified Plants, Messeweg 11/12, 38104 Braunschweig, Germany

Mark K. Sears, Department of Environmental Biology, University of Guelph, Guelph, Ontario N1G 2W1, Canada

Anthony M. Shelton, Department of Entomology, Cornell University/NYSAES, 630 W. North Street, Geneva, NY 14456, USA

Jeremy Sweet, Environmental Consultant, 6 The Green, Willingham, Cambridge CB24 5JA, UK

Zigfridas Vaituzis, Biopesticides and Pollution Prevention Division, USEPA 7511P, 1200 Pennsylvania Ave. N.W., Washington, DC 20460, USA

Jeffrey D. Wolt, Biosafety Institute for Genetically Modified Agricultural Products and Department of Agronomy, Iowa State University, 164 Seed Science Center, Ames, IA 50011, USA

References Canadian Food Inspection Agency (CFIA). 2004: Assessment Criteria for Determining

Environmental Safety of plants with Novel Traits. Directive 94-08. Canadian Food Inspection Agency, Government of Canada, Ottawa, Ontario.

Dutton, A., Romeis, J. & Bigler, F. 2003: Assessing the risks of insect resistant transgenic plants on entomophagous arthropods: Bt-maize expressing Cry1Ab as a case study. BioControl 48: 611-636.

European Community (EC). 2001: Directive 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC. Official Journal of the European Communities L 106: 1-39.

European Commission (EC). 2002: Commission Decision (2002/623/EC) of 24 July 2002 establishing guidance notes supplementing Annex II to Directive 2001/18/EC of the European Parliament and of the Council of the deliberate release in to the environment of genetically modified organisms and repealing Council Directive 90/220/EEC. Official Journal of the European Communities L 200: 22-33.

European Food Safety Authority (EFSA). 2006: Guidance document of the Scientific Panel on Genetically Modified Organisms for the risk assessment of genetically modified plants and derived food and feed. EFSA Journal 99: 1-100. http://www.efsa.europa.eu/EFSA/Scientific_Document/gmo_guidance_gm_plants_en,0.pdf

European Food Safety Authority (EFSA). 2007: Environmental Risk Assessment of Genetically Modified Plants - Challenges and Approaches. EFSA Scientific Colloquium No. 8. http://www.efsa.europa.eu/EFSA/ScientificOpinionPublicationReport/EFSAScientificColloquiumReports/efsa_locale-1178620753812_EnvironmentalRiskAssessmentofGeneticallyModifiedPlants.htm

155

Garcia-Alonso, M., Jacobs, E., Raybould, A., Nickson, T.E., Sowig, P., Willekens, H., van der Kouwe, P., Layton, R., Amijee, F., Fuentes, A.M. & Tencalla, F. 2006: A tiered system for assessing the risk of genetically modified plants to non-target organisms. Environmental Biosafety Research 5: 57-65.

James, C. 2006: Global Status of Commercialized Biotech/GM Crops: 2006. ISAAA Brief No. 35, International Service for the Acquisition of Agri-Biotech Applications, Ithaca, NY, USA.

Kuiper, H.A., Kleter, G.A., Noteborn, H.P.J.M. & Kok, E.J. 2001: Assessment of the food safety issues related to genetically modified foods. Plant Journal 27: 503-528.

Organisation for Economic Cooperation and Development (OECD). 1993: Safety considerations for biotechnology: scale-up of crop plants. Organisation for Economic Cooperation and Development, Paris. http://www.oecd.org/dataoecd/26/26/1958527.pdf?channelld=34537&homeChannelld=33703&fileTitle=Safety+Considerations+for+Biotechnology+Scale-up+of+Crop+Plants

Raybould, A. 2006: Problem formulation and hypothesis testing for environmental risk assessments of genetically modified crops. Environmental Biosafety Research 5: 119-125.

Romeis., J. 2004: Workshop report - Impact of GM crops on natural enemies. IOBC/WPRS Bulletin 27(3): 193-195.

Romeis, J. 2006: Non-target risk assessment of GM crops and regulation. IOBC/WPRS Bulletin 29(5): 197-200.

Romeis, J., Bartsch, D., Bigler, F., Candolfi, M.P., Gielkens, M.M.C., Hartley, S.E., Hellmich, R.L., Huesing, J.E., Jepson, P.C., Layton, R., Quemada, H., Raybould, A., Rose, R.I., Schiemann, J., Sears, M.K., Shelton, A.M., Sweet, J., Vaituzis, Z. & Wolt, J.D. 2006: Moving through the tiered and methodological framework for non-target arthropod risk assessment of transgenic insecticidal crops. Proceedings of the 9th International Symposium on the Biosafety of Genetically Modified Organisms, 24-29 September 2006, Jeju Island, South Korea, 62-67. http://www.isbr.info/symposia/

Romeis, J., Bartsch, D., Bigler, F., Candolfi, M.P., Gielkens, M.M.C., Hartley, S.E., Hellmich, R.L., Huesing, J.E., Jepson, P.C., Layton, R., Quemada, H., Raybould, A., Rose, R.I., Schiemann, J., Sears, M.K., Shelton, A.M., Sweet, J., Vaituzis, Z. & Wolt, J.D. 2007: Non-target arthropod risk assessment of insect-resistant GM crops. Proceedings of the V Brazilian Biosafety Congress and the V Latin American Symposium on Transgenic Products, Ouro Preto, Brazil, 18-21 September 2007, pp. 59-61.

Romeis, J., Bartsch, D., Bigler, F., Candolfi, M.P., Gielkens, M.M.C., Hartley, S.E., Hellmich, R.L., Huesing, J.E., Jepson, P.C., Layton, R., Quemada, H., Raybould, A., Rose, R.I., Schiemann, J., Sears, M.K., Shelton, A.M., Sweet, J., Vaituzis, Z. & Wolt, J.D., 2008b. Assessment of risk of insect-resistant transgenic crops to nontarget arthropods. Nature Biotechnology 26: 203-208.

Rose, R. (Ed.) 2007: White paper on tier-based testing for the effects of proteinaceous insecticidal plant-incorporated protectants on non-target arthropods for regulatory risk assessment. USDA-Aphis and US Environmental Protection Agency. http://www.epa.gov/pesticides/biopesticides/pips/non-target-arthropods.pdf

Touart, L.W. & Maciorowski, A.F. 1997: Information needs for pesticide registration in the United States. Ecological Applications 7: 1086-1093.

156

GMOs in Integrated Plant Production IOBC wprs Bulletin Vol. 33, 2008

pp. 157-158 Special activity: Non-target risk assessment and regulation. Protocol of the discussion Elisabeth Schulte Genius GmbH, Robert-Bosch-Str. 7, 64293 Darmstadt, Germany (E-mail: elisabeth.schulte@genius.de), Science Editor www.gmo-safety.eu The Non-target risk assessment and regulation initiative has been meeting regularly as part of the scientific IOBC/WPRS working group meetings on “GMOs in Integrated Plant Production” since 2005.

This year’s update presentation took place during the 3rd meeting in Warsaw, Poland from 23-25 May 2007. After an introduction by Jörg Romeis, coordinator of the initiative, Alan Raybould (Syngenta, UK), Franz Bigler (Agroscope ART, Switzerland) and Joe Huesing (Monsanto, USA) presented the developed non-target risk assessment approach. The presentations were then opened up for discussion by the 80 or more delegates. The moderator was Dr. Sabine Ebert, scientist and project leader (RWTH Aachen University, Institute of Environmental Research, Germany). For details on the presented approach, see Romeis (this volume) and Romeis et al. (2008; Nature Biotechnology 26, 203-208) Background The first genetically modified Bt maize was authorised in the USA over ten years ago. Now Bt maize and Bt cotton are grown on more than twenty million hectares worldwide. General guidance on conducting environmental risk assessments are available. However there is still a need for detailed descriptions of non-target species that need to be tested, and to establish test methods that apply to different regions.

The Non-target risk assessment and regulation initiative aims at establishing generic and pragmatic guidelines for assessing the risks of transgenic insecticidal crops for non-target organisms, particularly terrestrial non-target arthropods. The initiative involves scientists from diverse institutions including public research institutes, the agriculture biotech industry and representatives from regulatory agencies and a commercial testing laboratory.

The initiative has been meeting since 2005 within a small circle of experts and has presented its results for discussion at various symposia and expert discussions. Outcome discussion In many respects, the discussion round organised as part of the IOBC/WPRS meeting reflected the topic spectrum of past discussions (see http://www.gmo-safety.eu/en/maize/corn_borer/525.docu.html). The focus was on issues to do with methodology, especially the selection of test organisms for the laboratory experiments (surrogate species concept). The discussion concentrated on the following aspects: (1) Can a sufficient informative risk assessment that takes into account the diversity of ecosystems be carried out on the basis of a few representative organisms? (2) Are the selected surrogate species representative for the various cultivation regions?

157

158

158

(3) Is the methodological approach, derived from the pesticide authorisation methodology, helpful and does it lead to the desired results? (4) Should the methodological approach that focuses on the stressor, in this case the Bt protein, be extended to cover additional stressors, e.g. various climatic conditions? (5) Can the methodological approach also be applied to new insecticidal traits, non-insecticidal crops, such as pharmaceuticals, and stacked genes? Conclusion Targeted tests on representative organisms - the surrogate species concept There are 500–1000 species of arthropods per crop (resident and temporary) inhabiting our crops. A large number of them are not well known. Franz Bigler stressed the need to select appropriate species representing different ecological functions such as herbivory, pollination, predation and parasitism and decomposition in the soil for risk assessment. The selection of appropriate species to serve as surrogates for ecologically and economically important non-target arthropods is established and accepted in connection with pesticide authorisation. The surrogate species concept has also been adopted for the non-target arthropod risk assessment of Bt plants.

Regulatory testing of a chemical is not the same as a Bt protein, however the principles ofpesticide testing can be adopted, and test protocols have to be adapted. Differences mentioned in the discussion concerning the sensitivity of different species to the same test substance are taken into account by selecting the most sensitive representatives of a functional group for laboratory tests.

Proposals have been made for adequate species for most functional groups. Iidentification of adequate surrogates representing the soil ecosystem, however, still have to be developed. Problem formulation: defining the scope of the risk assessment Problem formulation is the very first step, essential to defining the scope of the risk assessment and the formulation of relevant risk hypotheses. In order to design adequate risk assessment studies, the specifics of the stressor, in this case the expressed trait, e.g. a Bt protein, are important. This includes the spectrum of activity and susceptibility, the mode of expression and the relevant exposure profiles.

The speakers pointed out that good problem formulation at the beginning of the risk assessment, taking into account the available experience and experiment results, leads to targeted laboratory tests that are designed in such a way that they can answer a large number of the questions relating to possible risks.

Franz Bigler believes that making the laboratory trials more complex by adding multiple stressors in addition to the insecticidal protein is not necessary and will not lead to more accurate results concerning the effect of the test compound. The methodological approach presented is pragmatic, generic and sufficiently informative.

The speakers believe their approach can also be used for new insecticidal traits, non-insecticidal crops, such as pharmaceuticals, and stacked genes.