Interciencia ISSN: 0378-1844 [email protected] Asociación Interciencia

Thomazoni, Danielle; Ferreira Soria, Miguel; Degrande, Paulo Eduardo; Faccenda, Odival; Jean Silvie, Pierre biodiversity index in Bollgard® cotton (Cry1Ac) in Interciencia, vol. 38, núm. 12, diciembre, 2013, pp. 849-856 Asociación Interciencia Caracas, Venezuela

Available in: http://www.redalyc.org/articulo.oa?id=33929617005

How to cite Complete issue Scientific Information System More information about this article Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal Journal's homepage in redalyc.org Non-profit academic project, developed under the open access initiative Arthropods biodiversity index in Bollgard® cotton (Cry1Ac) in Brazil

Danielle Thomazoni, Miguel Ferreira Soria, Paulo Eduardo Degrande, Odival Faccenda and Pierre Jean Silvie

SUMMARY

Shannon-Wiener’s diversity index (SWI) was used under un- get herbivores was significantly higher in Bt-cotton. The mean treated conditions of a cotton field during the 2006/2007 crop number of Anthonomus grandis (Boh.) and Edessa meditabun- season in the Cerrado region, Brazil. Comparison was carried da (Fabr.) adults were significantly higher in NuOpal® with the out between the transgenic NuOpal® (Bollgard®)(Cry1Ac) and whole plant sampling method. However, such differences were the non-transgenic isogenic variety DeltaOpal®. SWI was cal- not observed with the beat sheet method. For the natural ene- culated for target pests, non-target herbivores and predators mies, SWI and mean number of larvae and adults of the domi- groups. Two sampling methods were used: whole plant obser- nant predators did not show any significant difference between vation and beat sheet. As expected, the mean number of target Bt and non-Bt cotton. These results confirm the conservation pests, especially Pectinophora gossypiella (Saund.) and Ala- of some tritrophic interactions inside the Bt (untreated) cotton bama argillacea (Hübner), was significantly smaller in Bt cot- and contributes to a better sustainable management of non- ton. In the whole plant method sampling the SWI for non-tar- target pests by enhancement of their natural biological control.

Introduction vironment. The biological di- of existing species (Hurlbert, modified and the species have versity in one biological com- 1971). The importance of the to adapt to the modifications, Biological communities munity has two components: use of diversity indexes is so as to contribute with the have a degree of organization species richness (existing spe- their application in monitoring conservation of biodiversity in that is represented by their cies number) and homogenei- studies of biological commu- agroecosystems (Southwood, specific abundance distribu- ty, which depends on the nities dynamics and structural 1995). tion or relative frequency of larger or smaller uniformity change detection, when the Genetically modified (GM) the species present in the en- of the distribution frequency community environment is cotton varieties expressing the

KEYWORDS / Bt-cotton / Cry1Ac / Diversity / Herbivores Non-Target / Predators / Shannon-Wiener’s Index / Received: 25/09/2012. Modified: 09/12/2013. Accepted: 12/12/2013.

Danielle Thomazoni. Biologist Leste, MT, Brazil. 78850-000. Paulo Eduardo Degrande. Agron- Universidade Estadual de Mato and Ph.D. in Sciences (Ento- e-mail: daniellethomazoni@ omist and Ph.D. in Sciences Grosso do Sul, Brazil. mology), Universidade Federal imamt.com.br (Zoology), Universidade de São Pierre Jean Silvie. Biologist and de Grande Dourados (UFGD), Miguel Ferreira Soria. Agrono- Paulo, Brazil. Professor, UFGD, Ph.D. in Entomology, Centre de Brazil. Research Entomologist, mist and Ph.D. in Vegetal Pro- Brazil. Coopération Internationale en Instituto Mato-grossense do duction, UFGD, Brazil. Re- Odival Faccenda. Mathematician Recherche Agronomique Pour le Algodão (IMAmt), Brazil. Ad- search Entomologist, IMAmt, and Ph.D. in Agronomy (Energy Dévelopment (CIRAD), France. dress: Rodovia BR 070 Km Brazil. in Agriculture), Universidade Research Entomologist, CIRAD, 265, Campo Experimental, Estadual Paulista Júlio de Mes- Montpellier, France. Zona Rural, Primavera do quita Filho, Brazil. Professor,

DEC 2013, VOL. 38 Nº 12 0378-1844/13/12/849-08 $ 3.00/0 849 Índice de biodiversidad de artrópodOs en algodón Bollgard® (Cry1Ac) CULTIVADO en LA REGIÓN de cerrado en Brasil Danielle Thomazoni, Miguel Ferreira Soria, Paulo Eduardo Degrande, Odival Faccenda y Pierre Jean Silvie RESUMEN

El índice de biodiversidad de Shannon-Wiener’s (ISW) fue Bt. El ISW para los insectos plagas fue significativamente ma- utilizado en áreas dedicadas al cultivo de algodón no tratadas yor en algodón Bt con el método de planta entera, en tanto que insecticidas. El trabajo se realizó en áreas algodoneras locali- la media poblacional de adultos de Anthonomus grandis (Boh.) zadas en la región de Cerrado, Brasil. Se comparó el algodón y Edessa meditabunda (Fabr.) fue mayor en NuOpal® usando transgénico NuOpal® (Bollgard®)(Cry1Ac) con la isolínea no el mismo método. No obstante, esta diferencia no fue observa- transgénica DeltaOpal®. El índice de biodiversidad fue cal- da con el método de paño de sacudida. Por otra parte, el ISW, culado para todos los insectos presentes en el agroecosistema para los enemigos naturales y el valor de la media poblacional algodonero, incluyendo los insectos plagas de la variedad Bt, de y adultos de los predadores dominantes no presenta- convencionales y enemigos naturales. Los métodos de muestreo ron diferencia significativa entre el algodón Bt y el convencio- utilizados fueron el uso del paño de sacudida y la planta entera. nal. Estos resultados corroboran la conservación de las interac- Como era previsible, el valor promedio calculado de las plagas, ciones tritróficas en el algodón Bt (no tratado con insecticidas) específicamente Pectinophora gossypiella (Saund.) and Alabama y aporta nuevos elementos técnicos para manejo integrado de argillacea (Hübner), fue significativamente menor en el algodón insectos con énfasis en su control biológico natural.

índice de biodiversidade de artrópodes em algodão Bollgard® (Cry1Ac) no brasil Danielle Thomazoni, Miguel Ferreira Soria, Paulo Eduardo Degrande, Odival Faccenda e Pierre Jean Silvie RESUMO

O índice de diversidade de Shannon-Wiener (ISW) foi utili- maior em algodão Bt. O número médio de adultos de Antho- zado em condições de cultivo de algodoeiro não tratado, sem nomus grandis (Boh.) e Edessa meditabunda (Fabr.) foi signifi- aplicação de inseticidas durante a safra 2006/2007 no Cer- cativamente maior em NuOpal® utilizando o método de amos- rado, Brasil. Foi realizada comparação entre NuOpal® (Boll- tragem de planta inteira. Entretanto, esta mesma diferença não gard®)(Cry1Ac) e sua isolinha não transgênica. O índice foi foi observada com o método do pano de batida. Para inimi- calculado para pragas-alvo do algodão Bt, pragas não-alvo gos naturais, o ISW e o número médio de larvas e adultos de e inimigos naturais. Foram utilizados dois métodos de amos- predadores dominantes não apresentou diferença signficativa tragem: planta inteira e pano de batida. Como esperado, o entre algodão Bt e não-Bt. Estes resultados confirmam a con- número médio de pragas-alvo, especialmente Pectinophora servação de algumas interações tritróficas dentro do sistema gossypiella (Saund.) e Alabama argillacea (Hübner), foi signifi- algodão Bt (não pulverizado) e contribui para um manejo de cativamente menor em algodão Bt. Na amostragem por planta herbívoros não-alvo sustentável pelo incremento do seu contro- inteira, o ISW para herbívoros não-alvo foi significativamente le biológico natural.

Bacillus thuringiensis (Bt) and reduction of the risks to tween Bt and non-Bt cotton Bt and non-Bt cotton, consti- Cry1Ac protein (NuOpal® and human health (Shelton et al., was shown by Li et al. tuting a basis for the regula- DP90B) were introduced com- 2002; Naranjo, 2009). Another (2002), or an increase in the tion of population dynamics mercially in Brazil during the aspect is the promotion and communities diver- of pests and the dam- 2006/2007 crop season. the preservation of natural sity and pest sub-communi- age caused by such pests. Knowledge about the non- enemies, contributing to inte- ties (Men et al., 2003). In This paper presents the first target species (herbivores and grate pest management sys- Brazil, Ramiro and Faria study of biodiversity of non- natural enemies) present in tems with a strong biological (2006) observed no signifi- target herbivores and natural the Bt-cotton in different field control component and assess- cant differences in the total enemies (mainly predators) conditions is still incipient in ment of risk of Bt-cotton to predator specimens collected sampled with two methods: Latin America, in spite of the non-target arthropods, leading from Bollgard® cotton as whole plant and beat sheet, in economic importance of to a sustainable production compared to treatments with the Brazilian Cerrado Biome knowing the biological diver- and preserving the environ- Delta Pine Acala 90, with or (Savannah) region, Mato sity and maintaining a bio- ment (Romeis et al., 2006; without chemical control of Grosso do Sul State, Brazil. logical control during the in- 2008; Sarvjeet, 2012). caterpillars. A faunistic analysis of the troduction of GM crops (Ro- There are few published The objective of this re- genera and species found on meis et al., 2008; Lovei et al., studies conducted about the search was to study the ar- Bt-cotton compared with non 2009; Adenle, 2012). impact of the Bt-cotton vari- thropod biodiversity associat- Bt-cotton using the Shannon- Pest resistant GM varieties eties on the diversity of ar- ed with Bt-cotton (NuOpal®), Wiener’s index is discussed. were initially grown in coun- thropods, especially with re- as compared to the non-trans- tries such as the USA, Argen- spect to the values found in genic isogenic DeltaOpal® in Materials and Methods tina, Australia, China, diversity indexes, like the the absence of insecticide and South Africa, allowing Shannon-Wiener’s index. Us- sprays, promoting the knowl- This research was conduct- fewer insecticide applications, ing this index, no difference edge of structural changes in ed at the Faculdade de Ciên- reduction in production costs in arthropod biodiversity be- arthropod communities in the cias Agrárias (FCA), Univer-

850 DEC 2013, VOL. 38 Nº 12 sidade Federal da Grande Dou- pests, and natural enemies tion. Samplings for Heliothis the present study, a species rados, in Dourados, Mato found were performed every virescens and Alabama argil- was considered ‘dominant’ Grosso do Sul,, Brazil, be- seven days during the entire lacea eggs were made with the when its relative frequency tween February and June 2007. evaluation period, from the whole plant method. was >1/S, where S: total num- An irrigation system was in- crop’s VE stage (emergence), The diversity in Bt and non- ber of species found in the stalled in order to facilitate on February 21, until June 13. Bt cotton with both sampling sampling period. crop development during the Two visual sampling methods methods was based on calcula- In order to compare the experiment. Basic and side were used: ‘beat sheet’, recom- tions of frequency indices, con- mean differences between dressing fertilizations were mended by (Degrande et al., stancy, abundance, and domi- groups of target pests, non- performed with 400kg·ha‑1 2003), and ‘whole plant’. Sev- nance (Silveira-Neto et al., target pests, natural enemies NPK (8-20-20). Conventional enteen whole plant samplings 1976), considering the number and individuals of dominant planting system was adopted. and 8 beat sheet samplings of small (<1.0cm) and large species within each group, the Soil tillage was accomplished were made for each treatment, caterpillars (>1.0cm), larvae, development stage of the spec- on February 14, 2007, before preparing 272 replicates for the nymphs, and adults. Absolute imens was taken into consider- seeding by plowing and har- whole plant method and 128 frequency was defined as the ation. The comparison between rowing,. Seeds of the NuOpal® for the beat sheet method in total number of specimens ob- Bt and non-Bt treatments was (Bt) and DeltaOpal® (non-Bt) each treatment over the experi- served in the various sampling calculated based on the mean varieties used in the experi- mental period. conditions. of each treatment throughout ment were provided by MDM– In the beat sheet method, Constancy was defined as the entire sampling period. The Seeds of Cotton©, and were samplings were taken in the the percentage of samples in Student’s t test was later used pre-treated with the fungicides crop inter-row between the which a given species was at a significance level α=5%. Euparen® (tolylf luanid) three central rows of each sub present (Uramoto et al., 2005). The original data were not nor- (200g/100kg seeds), Monceren® area, at a random point, total- After the constancy percentag- mally distributed, and the test (pencycuron) (200g/100kg ing 16 points per treatment and es over the sampling periods was applied to the data trans- seeds), and Baytan® (triadime- observation date. The beat were obtained, the species formed to , thus meeting nol) (40ml/20Kg seeds), in or- sheet was white to facilitate were grouped into three cate- the assumptions associated der to control diseases that insect visualization; sheet gories: ‘constant’ (w), present with the model. cause damping-off. Both vari- width matched the crop row in more than 50% of the week- Target pest, non-target pest eties were manually seeded on spacing (0.9m), had a 1m ly observations; ‘accessory’ (y), and natural enemy diversity February 15, 2007, at a density length, and was adjusted so as present in 25 to 50% of the in Bt and non-Bt-cotton en- of 13 seeds/m and a row spac- to cover the inter-rows. Then, observations; and ‘accidental’ vironments were studied us- ing of 0.90m. Emergence oc- both rows were vigorously (z), present in less than 25% of ing Shannon-Wiener’s index curred 4 days later. Weed con- shaken, causing the , the observations. with a correction factor and trol was performed by hand- either immature or adult, to Abundance is the number of natural logarithm (Poole weeding during the entire cycle fall onto the sheet, allowing individuals of a given species 1974), by means of specimen for both varieties. Blitz® baits them to be visualized, counted, divided by the surface or vol- frequency. This index mea- (fipronil) were applied at the and identified at the family ume unit, and may vary in sures the degree of uncer- beginning of the crop cycle on and/or species level while still space and time (Silveira-Neto tainty in predicting to which the surroundings of the experi- in the field. The presence of et al., 1976). In order to esti- species will belong a ran- mental area to control leaf- the parasitoid Catolaccus gran- mate abundance, the limits domly selected individual, cutting ants of the genera Atta dis (Burks, 1954) and the tar- established by the confidence from a sample with S species and Acromyrmex. get lepidopteran Pectinophora interval (CI) at 5 and 1% and N individuals (Silveira- The total experimental area gossypiella was quantified in probabilities were used, and Neto et al., 1976). was 18×72m (0.12ha) with 32 this method by the number of the following five classes were The smaller the Shannon- sub-areas (plots) that were de- damaged reproductive struc- determined: ‘rare’ (r) with a Wiener’s index value, the marcated by random drawing, tures (bolls) fallen onto the number of individuals in the smaller the degree of uncer- 16 for each treatment: DeltaO- beat sheet. The bolls were then species smaller than the lower tainty, therefore reflecting the pal® and NuOpal®. Each Each opened to reveal individuals of CI limit at 1% probability; low diversity of a sample. sub-area comprised five rows those insects. ‘dispersed’ (d) with a number Diversity tends to be higher of the treatment variety, mea- In the whole plant evaluation of individuals between the for higher index values (Ura- suring 4.5×9m. The three cen- method, ten plants were evalu- lower limits of the confidence moto et al., 2005). Student’s t tral rows were sampled; one ated separately on the three intervals at 1% and 5% prob- test was used to check wheth- row at each end of each sub- central rows of each sub area, ability; ‘common’ (c) within er the species diversity differ- area was the border of the i.e., 160 plants per treatment the confidence interval at 5%; ence between those environ- sampling unit. In order to re- and observation date, by quan- ‘abundant’ (a), between the ments was significant at duce the incidence of Anthono- tifying and identifying the in- upper limits of the confidence α=5%. Data were analyzed mus grandis (Boheman, 1843) sects sampled at the family intervals at 5% and 1% prob- using the statistical software during the experiment, 23 and/or species level while still ability; and ‘very abundant’ package SPSS® (SPSS, 2006). traps were installed containing in the field. (va) with a number of indi- grandlure pheromone + insec- In both methods, when nec- viduals greater than the upper Results and Discussion ticide, to capture boll weevils essary, those insects that could CI limit at 1% probability. in the vicinity of the experi- not be identified in the field An organism is considered A total of 55 species were ment area. were collected and placed in a dominant when it receives observed, distributed among 11 The sampling and quantifica- recipient with 70% ethanol and impact from the environment orders and 32 families, and tion of the Bollgard® technolo- taken and after taken to the and becomes adapted to it were divided into three groups: gy target insects, non-target laboratory for later identifica- (Silveira-Neto et al., 1976). In target pests, non-target pests,

DEC 2013, VOL. 38 Nº 12 851 TABLE I Faunistic analysis of groups of target pests, non-target pests, and natural enemies by order, family, and species, sampling method, and type of cotton Sampling method Whole plant Beat sheet Group Order/Family Species Stage1 NuOpal® DeltaOpal® NuOpal® DeltaOpal® FC(A)D2 FC(A)D2 FC(A)D2 FC(A)D2 Lepidoptera/Noctuidae Alabama argillacea SC+LC 9 yn 55 w(c)s 0 31y(c)s Lepidoptera/Noctuidae Heliothis virescens SC+LC 0 1 z(c)n 0 1z(c)n pests Target Target Lepidoptera/Gelechiidae Pectinophora gossypiella Cat 0 39 y(c)s 0 35y(c)s Total 9 95 0 67 Coleoptera/Chrysomelidae Cerotoma arcuata Ad 5 z(d)n 2 z(d)n 0 0 Coleoptera/Chrysomelidae Chrysomelidae sp.1 Ad 2 z(d)n 2 z(d)n 0 1 z(d)n Coleoptera/Chrysomelidae speciosa Ad 52 w(c)n 45 w(c)n 35 w(c)s 26 w(c)n Coleoptera/Chrysomelidae Jansonius boggianii subaeneus Ad 37 w(c)n 38 w(c)n 13 y(c)n 21 y(c)n Coleoptera/Chrysomelidae Maecolaspis sp. Ad 19 y(d)n 10 y(d)n 0 0 Coleoptera/Cicindellidae Megascelis sp. Ad 1 z(d)n 2 z(d)n 0 1 z(d)n Coleoptera/Curculionidae Anthonomus grandis L+Ad 235 y(ma)s 154 y(c)s 47w(a)s 45 w(c)s Coleoptera/Lagriidae Lagria villosa Ad 2 z(d)n 3 z(d)n 8 w(d)n 8 y(d)n Coleoptera/Melyridae Astylus variegatus Ad 26 y(c)n 10 z(d)n 19 w(c)n 13 y(c)n /Aleyroidade Bemisia tabaci Ad 903 w(ma)s 856 w(ma)s 0 0 Hemiptera/ Neomegalotomus parvus Ad 0 3 z(d)n 1 w(r)n 1 z(d)n Hemiptera/Cicadellidae Agallia albidula Ad 315 w(ma)s 261 w(ma)s 11 w(c)n 12 z(c)n Hemiptera/Coreidae Hypselonotus sp. Ad 0 0 1 z(r)n 0 Hemiptera/Lygaeidae Oxycarenus sp. Ad 5 z(d)n 2 z(d)n 3 y(r)n 2 y(d)n Hemiptera/Miridae Horciasinus signoreti Ad 1 z(d)n 1 z(d)n 7 y(d)n 7 y(d)n Hemiptera/Miridae Horcias nobilellus Ad 611 w(ma)s 549 w(ma)s 142w(ma)s 150 w(ma)s Hemiptera/Pentatomidae Chinavia spp. N+Ad 5 z(d)n 2 z(d)n 6 y(d)n 2 y(d)n Non-target pests Hemiptera/Pentatomidae Edessa meditabunda N+Ad 129 w(c)s 74 w(c)n 61 w(ma)s 75 w(ma)s Hemiptera/Pentatomidae Euschistus heros N+Ad 114 w(c)s 93 w(c)s 40 w(c)s 31 w(c)s Hemiptera/Pentatomidae Nezara viridula N+Ad 63 w(c)n 61 w(c)n 40 w(c)s 31 w(c)s Hemiptera/Pentatomidae Piezodorus guildini N+Ad 38 w(c)n 17 w(c)n 6 w(d)n 6 y(d)n Hemiptera/Pyrrhocoridae Dysdercus sp. N+Ad 217 w(a)s 219w(ma)s 99 w(ma)s 165 w(ma)s Lepidoptera/Noctuidae Spodoptera eridania SC+LC 5 z(d)n 9 y(d)n 9 w(d)n 11 y(c)n Lepidoptera/Noctuidae Spodoptera frugiperda SC+LC 4 y(d)n 6 y(d)n 0 4 z(d)n Lepidoptera/Noctuidae Pseudoplusia includes SC+LC 4 y(d)n 13 y(d)n 7 w(d)n 12 w(c)n Orthoptera/Gryllidae Gryllus sp. Ad 3 z(d)n 4 z(d)n 0 1 z(d)n Orthoptera/Tettigoniidae Tettigoniidae sp.1 Ad 0 1 z(d)n 2 y(r)n 0 Thysanoptera/Thripidae Frankliniella sp. Ad 50 w(c)n 43 w(c)n 0 0 Total 2846 2480 557 625 Araneae Araneae Ad 152 w(va)s 185 w(va)s 58 w(va)s 72 w(va)s Coleoptera/Carabidae Callida sp. Ad 13 z(c)n 16 z(c)n 1 z(d)n 1 z(r)n Coleoptera/Carabidae Lebia concinna Ad 3 z(d)n 5 z(d)n 1 z(d)n 2 y(r)n Coleoptera/Coccinellidae Cycloneda sanguinea L+Ad 74 w(va)s 74 w(va)s 44 w(va)s 44 w(va)s Coleoptera/Coccinellidae Eriopsis connexa L+Ad 4 z(d)n 5 z(d)n 0 0 Coleoptera/Coccinellidae Hyperaspis festiva Ad 9 z(c)n 7 y(d)n 1 z(d)n 0 Coleoptera/Coccinellidae Olla v-nigrum Ad 3 z(d)n 5 z(d)n 2 z(d)n 1 z(r)n Coleoptera/Coccinellidae Scymnus sp. L+Ad 191 w(va)s 174 w(va)s 92 w(va)s 82 w(va)s Dermaptera/Forficulidae Doru luteipes Ad 7 y(c)n 12 y(c)n 4 y(d)n 7w(d)n Diptera/Dolichopodidae Condylostylus sp. Ad 1 z(d)n 1 z(d)n 0 0 Diptera/Syrphidae Toxomerus sp. L+Ad 7 y(c)n 8 y(d)n 1 z(d)n 0 Hemiptera/Anthocoridae Orius sp. Ad 27 y(c)n 45 y(c)s 23 y(c)s 41 y(a)s Hemiptera/Lygaeidae Geocoris sp. Ad 42 w(c)s 53 w(c)s 20 w(c)s 28 w(c)s Hemiptera/Nabidae Nabis sp. Ad 2 z(d)n 0 1 z(d)n 0 Hemiptera/Pentatomidae Podisus sp. N+Ad 2 z(d)n 3 z(d)n 3 y(d)n 5 y(d)n Natural enemies Natural Hemiptera/Reduviidae Repipta sp. Ad 1 z(d)n 2 z(d)n 0 0 Hemiptera/Reduviidae Zelus armillatus Ad 1 z(d)n 3 z(d)n 0 0 Hemiptera/Reduviidae Zelus longipes Ad 12 y(c)n 10 y(c)n 3 y(d)n 6 w(d)n Hymenoptera/Formicidae Solenopsis invicta Ad 25 y(c)n 17 y(c)n 1 z(d)n 1 z(r)n Hymenoptera/Pteromalidae Catolaccus grandis L 45 z(c)s 39 z(c)s 45 z(va)s 39 z(a)s Mantodea/Mantidae Mantidae sp.1 Ad 2 z(d)n 1 z(d)n 0 1 z(r)n Neuroptera/Chrysopidae Chrysoperla sp. L 28 y(c)s 21 y(c)n 26 w(c)s 20 w(c)n Neuroptera/Hemerobiidae Nusulala sp. L 2 z(d)n 8 y(d)n 2 y(d)n 5 y(d)n Neuroptera/Mantispidae Mantispidae sp.1 Ad 1 z(d)n 0 0 0 Total 654 694 328 355 Grand total 3509 3269 885 1047

1 SC: small caterpillar, LC: large caterpillar, Cat: caterpillar, L: , N: nymph, Ad: adult. 2 F: total number observed in different sampling conditions; C (constancy): w: constant, y: accessory, z: accidental; A (abundance): va: very abundant, a: abundant, c: common, d: dispersed, r: rare; D (dominance): s: dominant, n: non-dominant.

852 DEC 2013, VOL. 38 Nº 12 and natural enemies (Table I). Table II The very abundant (va) species Shannon-Wiener’s diversity index, (Variance), and number of non target herbivores, both of non-target pest species and natural enemies present in NuOpal® and DeltaOpal® in the Bt- and non-Bt cotton environments cotton with the whole plant Whole plant Bt cotton 1 (n=272) Non-Bt cotton 1 (n=272) t-Student P and beat sheet methods, were Non-target pests 2.11(0.004)(25) a 2.04(0.005)(27) b 1.98 0.047 Horciasoides nobilellus (Bergs- Natural enemies 2.17(0.002)(24) 2.21(0.001)(22) -0.558 0.576 ton, 1883) and Dysdercus sp. The species Bemisia tabaci Beat sheet Bt cotton 1 (n=128) Non-Bt cotton 1 (n=128) t-Student P (Gennadius, 1889) and Agallia Non-target pests 2.32(0.001)(20) 2.23(0.001)(22) 1.643 0.100 albidula (Uhler, 1895) were Natural enemies 2.04(0.002)(18) 2.12(0.002)(16) -1.053 0.292 very abundant in Bt and non- 1 Bt cotton only with the whole Different letters in a row represent non-significant values at 5%, assuming equal variances by Levene’s test. plant method. The species Edessa meditabunda (Fabr., (Fabr., 1798) and Dysdercus sp. Also, the sampling method stant in both Bt and non-Bt 1794), however, was very abun- were dominant in both variet- used can influence the quanti- cotton with the whole plant dant in both varieties, but only ies with both sampling meth- fication of each insect species sampling, and A. grandis and with the beat sheet method. ods, while B. tabaci and A. affecting the Bt and non-Bt Pseudoplusia includens (Walk- The non-target species An- albidula were dominant in both varieties, considering the be- er, 1857) were constant in both thonomus grandis (Boheman, NuOpal® and DeltaOpal® only havior of each species and types of cotton with beat sheet 1843) was found to be abun- with the whole plant method. their migration from sampling. The non-target herbi- dant in Bt cotton with both For the pentatomid species Ne- varieties to cotton (Lu et al., vores species Lagria villosa sampling methods. zara viridula (L., 1758) and E. 2010), as in the case of some (Fabr., 1783), Astylus variega- The fact of B. tabaci was meditabunda, the dominance pentatomids, searching food tus (Germar, 1824), Neomega- observed as the most abundant was observed in both varieties, resources in the cotton-soybean lotomus parvus (Westwood, species with similar population but only with the beat sheet agroecosystem. The correct 1842), A. albidula, Piezodorus densities between Bt and non-Bt method. In the whole plant selection of the sampling meth- guildini (Westwood, 1837) and cotton in the whole plant meth- sampling, the sucking herbi- od leads to real interpretations Spodoptera eridania (Cramer, od (Naranjo, 2005) can be at- vore species E. meditabunda about the effect of the GM 1782) were constant only in Bt tributed to the behavior of this was dominant in NuOpal® plant on the arthropod popula- cotton with the whole plant insect. Its quick flight when the only, while the chewing herbi- tion, and consequently on the method (Table I). plant is touched, made it diffi- vore Diabrotica speciosa (Ger- biological control, which may With regard to non-target cult to be sampled by beat mar, 1824) was dominant only be potentiated with the adop- herbivores diversity, the Shan- sheet, demonstrating the impor- in the Bt variety, with the beat tion of Bt crops. non-Wiener’s index with the tance of selecting the adequate sheet sampling (Table I). Considering the constancy whole plant method for Bt cot- sampling method to monitor These dominance results throughout the sampling peri- ton showed statistically signifi- non-target insects with an eco- showed the reduction in feed- od, A. argillacea had constant cant differences, being higher nomic importance in transgenic ing competition between non- incidence only in non-Bt cotton than non-Bt, thus demonstrat- varieties in the field (Naranjo et target herbivores and target with the whole plant method. ing that the NuOpal® variety al., 2005; Wade et al., 2006). insects controlled by the Boll- The constant non-target herbi- showed higher diversity of Another point that will be con- gard® technology, as a reduc- vore species in both NuOpal® non-target herbivores than the sidered is the differences in tion in competition for food and DeltaOpal® in both sam- DeltaOpal® variety with the ‘leaf hair’ between the cotton resources with target caterpil- plings were D. speciosa, H. whole plant sampling. Howev- varieties and in the mode of ac- lars controlled by the Cry1Ac nobilellus, E. meditabunda, E. er, in the beat sheet method, tion of Bt toxins inserted in toxin and the sucking herbi- heros, N. viridula (L., 1758) the index did not show signifi- these genetically modified cot- vores like the pentatomids and and Dysdercus sp., while B. cant difference (Table II). This ton varieties, which can influ- mirids. They show the impor- tabaci, A. albidula, Piezodorus result in the diversity index ence the abundance of insects, tance of the knowledge of bio- guildini (Westwood, 1837), can be explained by fact that as observed in Australia ecology interaction between Jansonius boggianii subaeneus the mean number of the non- (Whitehouse et al., 2007), with insects in an agroecosystem. and Frankliniella sp. were con- target herbivores A. grandis high numbers of whitefly in Bt cotton (VipCotton). Table III Dominance was found in the Mean number of arthropod specimens (SD) per type of cotton and target pest group for the spe- sampling method cies A. argillacea and P. gos- Whole plant Bt cotton 1 (n=272) Non-Bt cotton 1 (n=272) t-Student 2 p sypiella in non-Bt cotton, with both sampling methods. The Target pests 0.03 (0.19) a 0.35 (0.76) b 7.029 3 0.000 presence of the species H. vire- Non-target pests 10.45 (6.94) a 9.11 (6.68) b 2.626 0.009 scens was also detected, but at Natural enemies 2.36 (2.67) 2.50 (2.78) 0.817 0.414 a low frequency when com- Beat sheet Bt cotton 1 (n=128) Non-Bt cotton 1 (n=128) t-Student2 p pared with the other two target Target pests 0.00 (0.00) a 0.52 (0.96) b 6.777 3 0.000 pests, i.e., one caterpillar in 3 ® Non-target pests 4.35 (3.09) 4.87 (4.28) 0.498 0.619 DeltaOpal with both sampling Natural enemies 2.50 (2.38) 2.71 (2.21) 1.209 0.228 methods. The non-target herbi- vores species, A. grandis, H. 1 Different letters in a row represent non-significant values at 5%, assuming equal variances by Levene’s test. nobilellus, Euschistus heros 2 Original data transformed to for statistical analysis purposes. 3 Different variances.

DEC 2013, VOL. 38 Nº 12 853 Table IV individuals observed with both Mean number (SD) of individuals from dominant non-target pest sampling methods was signifi- species per type of cotton and sampling method cantly smaller in NuOpal® than in DeltaOpal®. Bt-cotton 3 Non-Bt cotton 3 Whole plant Stage 2 t-Student 1 p The results of the faunistic (n=272) (n=272) analysis of non-target species Horciasoides nobilellus Ad 2.25 (3.28) 2.02 (3.29) 0.848 0.397 of Cry1Ac (Bollgard® cotton) Anthonomus grandis Ad 0.86 (1.62) a 0.57 (1.19) b 2.267 0.024 sampled between the NuOpal® Agallia albidula Ad 1.16 (1.47) 0.96 (1.33) 1.703 0.089 and DeltaOpal® environments Nezara viridula N 0.06 (0.25) 0.03 (0.17) 1.345 0.179 demonstrate that abundance, Nezara viridula Ad 0.17 (0.59) 0.19 (0.56) 0.594 0.553 dominance and constancy of Euschistus heros N 0.04 (0.23) 0.03 (0.19) 0.787 0.431 Euschistus heros Ad 0.38 (0.84) 0.31 (0.95) 1.017 0.309 the species can be attributed to Edessa meditabunda N 0.04 (0.38) 0.03 (0.20) 0.319 0.750 several factors, such as the Edessa meditabunda Ad 0.43 (1.08) a 0.24 (0.73) b 2.451 0.015 lack of insecticidal activity of Dysdercus sp. N 0.06 (0.27) 0.07 (0.38) 0.316 0.752 the transgenic variety (Cry1Ac) Dysdercus sp. Ad 0.74 (1.41) 0.73 (1.33) 0.121 0.903 on non-target herbivores and ® Diabrotica speciosa Ad 0.19 (0.47) 0.17 (0.45) 0.679 0.498 predators in NuOpal , which can affect directly the diversity Bt-cotton 3 Non-Bt cotton 3 Beat sheet Stage 2 t-Student 1 p and trophic interactions of (n=128) (n=128) these insects, promoting the Horciasoides nobilellus Ad 1.11 (1.43) 1.17(1.52) 0.250 0.803 knowledge of these tritrophic Ad 0.37 (0.79) 0.35(0.78) 0.209 0.834 Anthonomus grandis interactions and the integration Agallia albidula Ad 0.09 (0.28) 0.09(0.50) 0.219 0.827 Nezara viridula N 0.08 (0.29) 0.06(0.27) 0.463 0.644 of Bt cotton use and biological Nezara viridula Ad 0.23 (0.55) 0.18(0.46) 0.812 0.418 control (Li et al., 2002; Romeis Euschistus heros N 0.00 (0.00) 0.02(0.12) 1.420 0.157 et al., 2008). Euschistus heros Ad 0.31 (0.64) 0.23(0.53) 1.111 0.268 Among the natural enemies Edessa meditabunda N 0.01 (0.08) 0.00(0.00) 1.000 0.318 sampled, predators were main- Edessa meditabunda Ad 0.47 (1.37) 0.59(1.97) 0.241 0.810 ly present. The predators Ara- Dysdercus sp. N 0.02 (0.15) 0.06(0.62) 0.438 0.662 neae, Cycloneda sanguinea (L., Dysdercus sp. Ad 0.75 (1.23) 1.23(2.03) 1.821 0.070 1763) and Scymnus sp. were Diabrotica speciosa Ad 0.27 (0.64) 0.20(0.49) 0.859 0.391 very abundant both in NuOpal® and DeltaOpal® with both sam- 1 2 3 Original data transformed to for statistical analysis purposes. N: nymph, Ad: adult. Different letters in a pling methods. With the beat row represent non-significant values at 5%, assuming equal variances by Levene’s test. sheet method, the parasitoid Table V species C. grandis was abun- dant both in NuOpal® and Del- Mean number of specimens (sd) of individuals from dominant natural ® enemy species per type of cotton and sampling method taOpal , and the predator bug Orius sp. was abundant in non Bt-cotton Non-Bt cotton Whole plant Stage 2 t-Student 1 P Bt cotton only. (n=272) (n=272) Dominant natural enemies Cycloneda sanguinea L 0.20 (0.58) 0.21 (0.82) 0.231 0.817 both in Bt and non-Bt cotton, Cycloneda sanguinea Ad 0.07 (0.31) 0.07 (0.29) 0.235 0.815 with both the whole plant and Scymnus sp. L 0.56 (1.29) 0.46 (1.15) 0.771 0.441 beat sheet methods, were Scymnus sp. Ad 0.15 (0.41) 0.18 (0.58) 0.425 0.671 Araneae, C. sanguinea, Scym- Chrysoperla sp. L 0.10 (0.51) 0.08 (0.28) 0.332 0.740 nus sp., Geocoris sp. and C. Geocoris sp. Ad 0.15 (0.44) 0.19 (0.53) 0.951 0.342 grandis. On the other hand, Orius sp. Ad 0.10 (0.45) 0.17 (0.62) 1.326 0.185 the predator Orius sp. was Araneae Ad 0.56 (0.93) 0.68 (1.00) 1.629 0.104 dominant in both NuOpal® Catolaccus grandis L 0.14 (0.62) 0.12 (0.64) 0.533 0.594 and DeltaOpal® with the beat 2 Bt-cotton Non-Bt cotton sheet method. However, this Beat sheet Stage t-Student 1 P (n=128) (n=128) bug was only dominant in Cycloneda sanguinea L 0.27 (0.70) 0.27 (1.09) 0.577 0.564 DeltaOpal® with the whole Cycloneda sanguinea Ad 0.07 (0.31) 0.08 (0.26) 0.367 0.714 plant sampling. The beneficial Scymnus sp. L 0.61 (1.13) 0.56 (0.91) 0.063 0.950 arthropods Araneae, C. san- Scymnus sp. Ad 0.11 (0.36) 0.08 (0.26) 0.694 0.489 guinea, Scymnus sp. and Geo- Chrysoperla sp. L 0.20 (0.71) 0.16 (0.38) 0.258 0.797 coris sp. Did not vary in both Geocoris sp. Ad 0.16 (0.40) 0.22 (0.46) 1.160 0.247 Bt and non-Bt cotton with Orius sp. Ad 0.18 (0.63) 0.32 (0.84) 1.518 0.130 both sampling methods. How- Araneae Ad 0.45 (0.85) 0.56 (0.81) 1.343 0.180 ever, in the beat sheet, the Catolaccus grandis L 0.29 (0.88) 0.25 (0.93) 0.546 0.585 predators Doru luteipes 1 Original data transformed to for statistical analysis purposes. 2 L: larva, Ad: adult. (Scudder, 1876) and Zelus longipes (L., 1767) were con- stant in DeltaOpal® only, and E. meditabunda adults in Bt and non-Bt cotton, being was not observed with the beat while Chrysoperla sp. was the whole plant method was higher in Bt cotton than non-Bt sheet method (Table IV). The constant in both Bt and non- significantly different between (Table III). Yet, such difference mean number of target pest Bt cotton (Table I).

854 DEC 2013, VOL. 38 Nº 12 The faunistic analysis for cal control application during system in Brazil, being rele- non-target arthropod popula- the natural enemies is mainly the entire development cycle of vant because the arthropods tions in Brazil. correlated with the tritrophic both varieties can influence play an important part in the interactions and the direct and predator abundance, and influ- structure and operation of the Acknowledgements indirect effects of Bt toxin enced this genus, which was ecosystems and the mainte- plants on the preys and hosts abundant in both Bt and non- nance of the biological diver- The authors thank Sérgio of these benefical insects Bt cotton. This result was con- sity (Tscharntke and Clough, Vanin (Universidade de São (Shelton et al., 2002). Another firmed by the measure of the 2007; Scherr and McNeely, Paulo) for identifying the point to be considered is the mean number of individuals of 2008; Sarvjeet, 2012). With the chrysomelid Jansonius bog- traits adopted in the trans- Orius sp., and the same result crescent introduction of the Bt gianii subaeneus, Rogério genic varieties (Wan et al., was observed with Araneae in cotton varieties in Brazil, it is Silvestre and Manoel Fernan- 2002; Yang et al., 2005), like both sampling methods. But crucial to quantify the diversity do Demétrio (Universidade the application or not of insec- other studies have shown a of insect communities present Federal da Grande Dourados) ticides with regard to the con- higher mean number of these in Bt and non-Bt cotton plots for identifying the ants (Sole- trol level to the non-target predators in Bt cotton than in and to determine how these nopsis invicta e Atta sp.), herbivores of Bt plants (De- non Bt cotton (Wan et al., communities are influenced by MDM-Sementes de Algodão© grande, 2004; Thomazoni et 2002; Hagerty et al., 2005). environmental changes pro- for the seeds of both varieties al., 2010), the use of selective The similarity in the mean voked by natural causes or by used in the experiment and insecticides, and the sampling number of individuals of the human activity, like agronomic CNPq for the grant awarded method adopted correlated Chrysopidae family between Bt practices such a insecticide to the first author. with the behavior and bioecol- and non-Bt cotton was also control for non-herbivores in- ogy of the natural enemy, observed in Australia (White- sects, and how these affect the References which can be influence the house et al., 2005), and the biodiversity by tritrophic inter- diversity of these insects (Men presence of individuals of this actions that can contribute with Adenle AA (2012) Failure to achieve et al., 2003), as demonstrated family was also observed in Bt the reduction of target pests, as 2010 biodiversity’s target in de- veloping countries: How can in this study, in the beat sheet cotton (Sisterson et al., 2004). natural enemies are the main conservation help? Biodiv. Cons. method for Orius sp., which However, Hagerty et al. (2005) cause of insect mortality in 21: 2435-2442. was dominant in both NuO- found that chrysopid popula- agroecosystems (Parra, 2000; Degrande PE, Oliveira MA, Ribeiro pal® and DeltaOpal® and was tions belonging to the Peixoto et al., 2007). In this JF, Barros R, Nogueira RF, Ro- only dominant in DeltaOpal® Chrysopidae and Hemerobiidae way, this research can show drigues ALL, Fernandes MG with the whole plant sampling. families were more abundant how to integrate natural con- (2003) Avaliação de métodos ® para quantificar predadores de With regard to natural ene- in Bollgard cotton when com- trol with the transgenic plants, pragas do algodoeiro. Arq. Inst. mies diversity in both sam- pared with non-Bt cotton, promoting the conservation of Biol.. 70: 291-294. pling methods, the Shannon- showing a negative effect of Bt the beneficial insects, using the Degrande PE (2004) Níveis de con- Wiener’s index and also the cotton on the diversity of these different monitoring sampling trole das pragas do algodoeiro. mean number of natural ene- natural enemies. and diversity indexes. Atual. Agric. 1: 22-23. mies was not significantly dif- A lack of significant differ- Hagerty AM, Kilpatrick AL, Turnip- ferent between cotton varieties ence in the mean number of Conclusions seed SG, Sullivan MJ (2005) using both sampling methods sampled Coccinellidae was also Predaceous arthropods and lepi- ® dopteran pests on conventional, (Table II). The mean number observed in China (Yang et al., NuOpal (Cry1Ac) is effi- Bollgard, and Bollgard II cotton of larvae and adults in the 2005). However, in other coun- cient in the control of target under untreated and disrupted dominant genera and species tries (Hagerty et al., 2005; species (P. gossypiella, H. vire- conditions. Env. Entomol. of the predators C. sanguinea, Hofs et al., 2005) differences scens and A. argillacea) under 34(5):105-114. Scymnus sp., Chrysoperla sp., in the mean number of speci- cultivation conditions of the Hilbeck A, Andow DA, Arpaia S, Geocoris sp., Orius sp., Ara- mens from this family of pred- Brazilian Cerrado biome (sa- Birch ANE, Fontes EMG, Lövei GL, Sujii ER, Wheatley RE, neae and C. grandis did not ators have been observed. Such vannah) without insecticide Underwood E (2006) Methodol- show significant differences difference in results can be at- sprays. ogy to support non-target and between Bt and non-Bt cotton tributed to the number of prey The whole plant sampling biodiversity risk assessment. In in any of the sampling meth- sampled between Bt and non- method, as detected by the di- Hilbeck A, Andow DA, Fontes ods (Table V). Bt cotton, whose development versity indexes, has a higher EMG (Eds.) Environmental Risk Assessment of Genetically Mod- The dominance of predator was favored by the lack of ac- diversity of the non-target her- ified Organisms: Methodologies Chrysoperla sp. in Bt cotton tion of insecticides (Marvier et bivores A. grandis and E. med- for Assessing Bt Cotton in Bra- with both sampling methods, al., 2007) that are otherwise itabunda in Bt cotton (Cry- zil. 1st ed. CABI. Cambridge, may indicate that this insect commonly applied for their 1Ac). MA, USA. pp. 108-132. possibly did not suffer a nega- control, showing the impor- The natural enemies diversi- Hofs JL, Schoeman A, Mellet M, tive impact from the Bt toxin tance of integrating pest man- ty on the non-sprayed Bt cot- Vaissayre M (2005) Impact des cotonniers génétiquement modi- present in the transgenic vari- agement tactics (Romeis et al., ton (Cry1Ac) shows tritrophic fiés sur la biodiversité de la ety (Hilbeck et al., 2006). In 2006, 2008), in this case, Bt interactions, and the conserva- faune entomologique: Le cas du contrast to our results, a small cotton and biological control. tion potential and benefits on coton Bt em Afrique du Sud. difference in Geocoris sp. pop- This biodiversity study was that agroecossystem. Int. J. Trop. Ins. Sci. 25: 63-72. ulation density between Bt and conducted to better understand Moreover, this research Hurlbert SH (1971) The non-concept non-Bt cotton, both without the biology and ecology of the demonstrated that the faunis- of species diversity: a critique and alternative parameters. Eco- chemical control application, predator/pest interactions in Bt tic analysis and the diversity logy 52: 577-586. was observed by Naranjo and non-Bt cotton varieties index of Shannon-Wiener can Li WD, Wu KM, Chen XX, Feng (2005). In the case of Orius sp. without application of insecti- be used in studies of risk as- HQ, Guang X, Guo YY (2002) populations, the lack of chemi- cides, in a savannah agroeco- sessment of Bt varieties in Effects of transgenic cotton car-

DEC 2013, VOL. 38 Nº 12 855 rying Cry1A+CpTI and Cry1Ac terns. CAB Reviews: Persp. sinenea E (Ed,) Bacillus of arthropods under practical genes on diversity of arthropod Agric. Vet. Sci. Nut. Nat. Res. thuringiensis Biotechnology. field conditions in Brazil. Afr. J. communities in cotton fields in 4: 1-23. Springer. New York, USA. pp Biotechnol. 9(37): 6167-6176. North China. Chin. J. Agric. Parra JRP (2000) O controle bi- 41-85. Uramoto K, Walder JMM, Zucchi Biotechnol. 1: 17-21. ológico e o manejo de pragas: Scherr SJ, McNeely JA (2008) Bio- RA (2005) Análise quantitativa Lövei GL, Andow DA, Arpaia S passado, presente e futuro. In diversity conservation and ag- e distribuição de populações de (2009) Transgenic insecticidal Guedes JC, Costa ID, Casti- ricultural sustainability: to- espécies de Anastrepha (Dip- crops and natural enemies: a glioni E (Eds.) Bases Técnicas wards a new paradigm of ‘eco- tera: Tephritidae) no Campus detailed review of laboratory do Manejo de Insetos. UFSM. agriculture’ landscapes. Phil. Luiz de Queiroz, Piracicaba, SP. studies. Env. Entomol. 38: CCR.DFS. Santa Maria, Brazil. Trans. Roy. Soc. Lond. B. 363: Neotrop. Entomol. 34: 33-39. 477-494. 293-306. pp. 59-70. Wade MR, Scholz BCG, Lloyd RJ, Lu Y, Wu K, Jiang Y, Xia B, Li P, Peixoto MS, Barros LC, Bassanezi Shelton AM, Zhao JZ, Roush RT Cleary AJ, Franzmann BA, Feng H, Wyckhuys KAG, Guo RC (2007) Uma abordagem (2002) Economic, ecological, Zalucki MP (2006) Temporal Y (2010) Mirid bug outbreaks fuzzy para um modelo presa- food safety, and social conse- variation in arthropod sampling in multiple crops correlated predador acoplado ao parasitis- quences of the deployment of Bt effectiveness: the case for using with wide-scale adoption of Bt mo. Tema - Tend. Mat. Aplic. transgenic plants. Annu. Rev. the beat sheet method in cot- cotton in China. Science 328: Comput. 8: 119-128. Entomol. 47: 845-881. ton. Entomol. Exp. Appl. 120: 1151-1154. Poole RW (1974) An Introduction to Southwood TRE. (1995) Ecological 139-153. Marvier M, McCreedy C, Regetz J, Quantitative Ecology. McGraw- Methods: With Particular Refer- Wan, FH, Liu WX, Guo JY (2002) Kareiva P (2007) A meta-analy- Hill. New York, USA. ence to the Study of Insect Pop- Comparison analyses of the ulations. Chapman & Hall. Lon- sis of effects of Bt cotton and Ramiro ZA, Faria AM (2006) Le- functional groups of natural en- don, UK. 524 pp. on nontarget inverte- vantamento de insetos preda- emy in transgenic Bt-cotton field brates. Science 316: 1475-1477. dores nos cultivares de algodão Silveira Neto S, Nakano O, Barbin and non-transgenic cotton fields Men X, Ge F, Liu X, Yardim EN Bollgard®DP90 e convencional D, Vila Nova NA (1976) Manu- with IPM and chemical control. (2003) Diversity of arthropod Delta Pine Acala 90. Arq. Inst. al de Ecologia dos Insetos. Ce- Acta Ecol. Sin. 22: 935-942. communities in transgenic Bt Biol. 73: 119-121. res. São Paulo, Brazil. 419 pp. Whitehouse MEA, Wilson LJ, Fitt cotton and nontransgenic cotton Romeis J, Meissle M, Bigler F Sisterson MS, Biggs RW, Olson C, GP (2005) A comparison of ar- agroecosystems. Env. Entomol. (2006) Transgenic crops ex- Carriére Y, Dennehy TJ, Ta- thropod communities in trans- 32: 270-275. pressing Bacillus thuringiensis bashnik BE (2004) Arthropod genic Bt and conventional cot- Naranjo SE (2005) Long-term as- toxins and biological control. abundance and diversity in Bt ton in Australia. Env. Entomol. sessment of the effects of trans- Nat. Biotechnol. 24: 63-71. and Non-Bt cotton fields. Env. 34: 1224-1241. Entomol. 33: 921-929. genic Bt-cotton on the abun- Romeis J, Bartsch D, Bigler F, Can- Whitehouse MEA, Wilson LJ, Con- dance of nontarget arthropod dolfi M, Gielkens MM, Hartley SSPS (2006) Statistical Package for stable GA (2007) Target and natural enemies. Env. Entomol. SE, Hellmich RL, Huesing JE, Social Sciences 15.0. SPSS Inc. non-target effects on the inver- 34: 1193-1210. Jepson PC, Layton R, Quemada Chicago, Il, USA. 179 pp. tebrate community of Vip cot- Naranjo SE, Head G, Dively GP H, Raybould A, Rose RI, Shie- Tscharntke T, Bommarco R, Clough ton, a new insecticidal trans- (2005) Field studies assess- mann J, Sears MK, Shelton Y (2007) Conservation biologi- genic. Aust. J. Agr. Res. 58: ing arthropod nontarget ef- AM, Sweet J, Vaituzis Z, Wolt cal control and enemy diversity 273-285. fects in Bt transgenic crops: JD (2008) Assessment of risk of on a landscape scale. Biol. Con- Yang P, Li K, Shi S, Xia J, Guo R, Li Introduction. Env. Entomol. insect-resistant transgenic crops trol 43: 294-309. S, Wang L (2005) Impacts of 34: 1178-1180. to nontarget arthropods. Nat. Thomazoni D, Degrande PE, Silvie transgenic Bt cotton and integrat- Naranjo SE (2009) Impacts of Bt Biotechnol. 26: 203-208. PJ, Faccenda O. (2010) Impact ed pest management education on crops on non-target inverte- Sarvjeet K (2012) Risk assessment of Bollgard® genetically modi- smallholder cotton farmers. Int. J. brates and insecticide use pat- of Bt transgenic crops. In San- fied cotton on the biodiversity Pest Manag. 51: 231-234.

856 DEC 2013, VOL. 38 Nº 12