Postcommercialization Monitoring of the Long-Term Impact of BtCorn on Non-Target Arthropod Communities in Commercial Farms

and Adjacent Riparian Areas in the Philippines

EDWIN P. ALCANTARA1

National Institute of Molecular Biology and Biotechnology, University of the Philippines Los Banos,College, Laguna 4031, Philippines

Environ. Entomol. 41(5): 1268Ð1276 (2012); DOI: http://dx.doi.org/10.1603/EN10297

ABSTRACT A 4-yr on-farm study was conducted in the Philippines between 2006 and 2009 todetermine if there are any long-term impacts of Bt corn on the arthropod community on commercialfarms and in adjacent riparian areas. Arthropod counts were gathered by visual inspection of cornplants in three pairs of commercial farms and by sweep sampling in riparian sites close by. Samplingshowed that species composition between Bt and non-Bt corn and between riparian areas adjacentto Bt or non-Bt corn were similar. Principal response curves and analysis of variance showed that therewere no adverse effects of Bt corn on the abundance of natural enemies either in crops or adjacentriparian sites. The frequently observed natural enemies in Bt corn were the predatory coccinellidbeetle, Micraspis discolor (F.) and spiders (Araneae). Arthropod diversity was similar 1) in Bt- andnon-Bt corn and 2) the adjacent riparian areas. Taken together, the two parts of this study show thatBt corn does not have any long-term adverse effect on arthropod communities in corn (Zea mays L.)or in adjacent riparian ecosystems. The methods and results described in this study will be useful toPhilippine government regulators in making recommendations and policy guidelines related to mon-itoring for environmental effects of transgenic crops. The current study did not attempt to demonstratethe direct exposure of arthropods in riparian sites to transgenic Cry1Ab protein that might come fromadjacent Bt corn Þelds; that would require different experimental methods but the current studysuggests such effects are unlikely.

KEY WORDS monitoring, Bt corn, arthropod abundance, PRC analysis, riparian area

Bacillus thuringiensis Berliner (Bt) is a gram-positivebacterium that produces insecticidal Cry proteins.Through genetic engineering, several crops express-ing Cry proteins have acquired protection from dam-age because of insect attack, including the Bt corncontaining Cry1Ab that is currently sold in the Phil-ippines. Since the governmentÕs approval of Bt-corncommercial propagation in December of 2002, thePhilippines has become one of the worldÕs major usersof the Bt-corn technology, with the land area plantedwith Bt corn reaching 600,000 ha at the end of2011 (James 2011). The adoption of Bt corn hassigniÞcantly increased corn-farmer productivity(Yorobe and Quicoy 2006, Brookes and Barfoot 2008)because of effective control of Ostrinia furnacalis(Guenee), and possibly also because of the positiveeffect of the associated regulatory system and researchcapability (Raney 2006).

BeneÞcial natural enemies have an important role inregulating the population of insect pests in corn (Zeamays L.) and monitoring for their abundance in theÞeld constitutes a valid risk-assessment methodology

for assessing the impacts of this new technology(Romeis et al. 2010). Natural enemies are exposed totransgenic Cry1Ab protein in Bt corn either throughbitrophic or tritrophic interactions (Naranjo 2009).Another environmental-risk concern is the exposureof any adjacent riparian arthropod community totransgenic Cry1Ab protein transported by mobile ar-thropods, corn pollen, and detritus (Rosi-Marshall etal. 2008) from the transgenic corn Þeld and potentiallyaffecting overall biodiversity. This has been consid-ered in various risk assessments (GrifÞths et al. 2009,Jensen et al. 2010, Tank et al. 2010) because riparianareas are also important for the conservation of ar-thropod species diversity. However, it is generallyconsidered that the risk of exposure to transgenic Cryproteins is very low (Volkmar et al. 2004, Storer et al.2008, Jensen et al. 2010, Tian et al. 2010). The currentstudy addresses the potential long-term impact fromexposure of arthropods to transgenic Cry1Ab proteinexpressed in Bt corn, both inside the transgenic-cornecosystem and in adjacent riparian areas. The speciÞcobjective of the current study was to monitor thelong-term effect of Bt corn on arthropod communitycomposition and diversity, including natural enemies,1 Corresponding author, e-mail: epalcantara@uplb.edu.ph.

Reproduced from Environ. Entomol. 41: 1268-1276 (2012).

Edwin P. Alcantara: Participant of the 14th UM, 1986-1987.275

between Bt- and non-Bt corn and in adjacent riparianecosystems.

Materials and Methods

Experimental Design. Only a limited number ofsiteswithadjacent riparianareaswereavailable for thecurrent study because relatively few owners with suit-able sites were willing to volunteer their farms. How-ever, six suitable farms were available at Ocampo,Sangay, and Tigaon in the province of Camarines Sur,the Philippines, and so the paired-Þeld design wasreplicated three times in each of Þve cropping seasonsover 4 yr, as is appropriate for Tier 4 studies (Perry etal. 2009). The cropping seasons 1Ð5 spanned JuneÐSeptember 2006, JanuaryÐApril 2007, JuneÐSeptember2007, JuneÐSeptember 2008, and JanuaryÐApril 2009,respectively. The experiment was carried out as arandomized complete-block design by using the threepairs of commercial farms, with each pair of farmsserving as a replicate. The farms in each pair were �1ha each, no �1 km from each other and with theadjacent riparian areas all �25 m from the corn Þelds.Crop type, speciÞcally Bt corn ÔDekalb 818 YieldGardÕand near isoline non-Bt corn ÔDekalb 818�, served asthe treatments. Each crop type was planted in thesame 1-ha Þeld each season on the same farms. Farmerpractice in all corn Þelds included the application ofthe insecticide Larvin (Bayer CropScience, ResearchTriangle Park, NC) at planting for the control ofAther-igona oryzae Malloch.Sampling. Arthropod counts were determined by

visual sampling as described by Dively (2005), withmodiÞcations. On each farm, 100 corn plants weresampled in a “W” pattern that covered the whole farm;this was similar to the sampling protocol used by Can-dolÞ et al. (2004). All above-ground parts of each plantwere examined carefully for arthropods, includingboth sides of leaves. Arthropod specimens collectedfrom visual samplings were identiÞed to species level,whenever possible. All sampling activities were con-ducted early in the morning when arthropods wereless active. Sampling occurred 25, 35, 45, 55, 60, and110 d after planting, which corresponded with theearly vegetative, middle vegetative, late vegetative,reproductive tasseling, reproductive silking, and ma-turity stage, respectively. Sweep sampling was con-ducted in riparian sites. In each riparian site, sweepsampling was carried out in each of Þve 5-m2 quadrats.The distance to the edge of each farm to the ripariansite was variable and ranged from 5 to �25 m. Fortysweeps were collected in each quadrat. All arthropodcollections from in each quadrat were preserved in70% alcohol and identiÞed in the laboratory. Samplingin the riparian sites coincided with the farm sampling.Statistical Analysis.Principal response curve (PRC)

analysis is a community-based multivariate methoddesigned to test and display treatment effects thatchange across time (Van den Brink and Ter Braak1999). The principal response can be explained as aÔvirtual speciesÕ that reßects the behavior of the wholecommunity (CandolÞ et al. 2004). Each set of season-

al-abundance data gathered from June 2006 to April2009 was separately analyzed by PRC to measure thelong-term impact of Bt corn on arthropod-communitycomposition. Abundance data were �(x � 1)-trans-formed before analysis to normalize the data (Jong-man et al. 1995). SigniÞcant deviation of the principal-response canonical coefÞcient (Cdt) for Bt corn fromcontrol (i.e., near isoline non-Bt corn) over all sam-pling dates was tested using the Monte-Carlo method(999 permutations) performed with Canoco 4.5 (TerBraak and Smilauer 2002). Each pair of farms wasconsidered as a block as deÞned by covariables. Sam-pling time was treated as a covariate and the productof sampling time and crop type (i.e., Bt- and non-Btcorn) as explanatory variables. Sampling of split plots(i.e., sampling dates) was restricted, whereas wholeplots (treatments) were shufßed randomly. The effectof crop type on abundance of major arthropod groupsor species was determined by repeated-measures anal-ysis of variance (ANOVA) with repeated measures onsampling dates within season. Data were transformedonto the �(x � 0.5) scale before statistical analysis byusing SYSTAT version 12. The same statistical treat-ments were applied to riparian abundance data. Thediversity index, eH (i.e., the exponential of Shannonindex), was used to estimate arthropod diversity (Jost2006, Fraulo et al. 2008) in both corn and riparian sitesas this provides a more useful comparative measure ofdiversity than the raw Shannon Index scores (Jost2006). All taxa within each data set were used in theanalyses.Repeated-measuresANOVAalsowasused totest the null hypothesis that there is no statisticallysigniÞcant difference in diversity index between Bt-and non-Bt corn and between riparian sites adjacentto Bt- and non-Bt corn farms. Where there were sig-niÞcant treatment � sampling date interactions, treat-ment differences were tested using the Protected leastsigniÞcant difference test with a Bonferroni probabil-ity correction to maintain a family-wise probabilitylevel of P 0.05.

Results

ArthropodCommunity Inside theCornEcosystem.SigniÞcant community-level impact of Bt corn on ar-thropod fauna was not detected by PRC analysis in anyof the Þve seasons, as shown in Fig. 1. Crop typeaccounted for only 11.1, 10.4, 6.9, 16.5, and 6.1% of thetotal variance for seasons 1Ð5, respectively, of which40.2, 42.4, 35.1, 52.9, and 34.5%, respectively, werecaptured by the Þrst PRC axis (plotted as the Y-axis inFig. 1). A large proportion of variance in the commu-nities was explained by changes during the seasonbecause sampling dates accounted for 45.0, 35.6, 58.0,47.0, and 60.6% of the total variance in the communi-ties over the Þve cropping seasons and blocking ac-counted for only 8.0, 8.0, 7.0, 5.0, and 8.5% of the totalvariance. In the Þgures, taxa with weights between0.5 and �0.5 were not shown because they are likelyto showaweakresponseora response that isunrelatedto the PRC curve (Van den Brink and Ter Braak 1999).The species weights for each taxon shown on the right

276

side of the PRC curves correspond to the weight ofeach species for the response shown in the respectivePRC plot. Taxa with positive contributions to the PRCbehave similarly to the principal response, i.e., if theprincipal response is increasing in a treated group, thetaxa with positive contributions tend to increase aswell. However, taxa with negative contributions tendto decrease if the principal response curve is increas-ing (CandolÞ et al. 2004). The natural enemies thatwere shown to have a strong afÞnity with the overallcommunity response as displayed in the PRC diagramswere spiders (Araneae), the coccinellid beetles Mi-craspisdiscolor(F.) andChilomenes sexmaculatus(F.),tropical Þre ant Solenopsis geminata (F.), a group ofunidentiÞed parasitic Hymenoptera including Ich-neumonidae, lace wings (Chrysopidae), and Syrphi-dae.

The average seasonal abundance of these naturalenemies is shown in Table 1. For all analyses exceptone (M. discolor, season 4), there was no treatment �time interaction and so time and treatment effectswere considered at the main effect level only. Pairwisetests of M. discolor numbers for each sampling dateduring season 4 showed that there were no signiÞcantdifferences inM. discolor abundance between the Bt-and non-Bt corn on any sampling date in that season.For all other natural enemies, there was no signiÞcant

difference between Bt- and non-Bt corn at the main-effect level in any season (Table 1). In addition, croptype had no signiÞcant inßuence on the community-diversity index (Table 1).Arthropod Community Inside the Riparian Area.

There were no signiÞcant differences from the control(non-Bt) in thePRCcurvesobtained fromtheripariansites adjacent to Bt corn (Fig. 2). The adjacent croptype accounted for only 10.1, 13.9, 7.4, 13.7, and 8.8%of the variance for the Þve seasons, of which 27.6, 39.6,30.7, 44.0, and 37.4%, respectively, of this variance wascaptured by the Þrst PRC axis (Fig. 2, Y-axis). A largeproportion of variance in the communities was ex-plained by changes during the season because sam-pling dates accounted for 34.1, 32.3, 45.2, 36.1, and55.3% of the total variance in the communities in eachof the Þve seasons, whereas blocking accounted for10.2, 7.7, 7.1, 8.5, and 10.9% of the variance. The naturalenemies that had a strong afÞnity with the overallcommunity response (Fig. 2) were dragonßies(Anisoptera), damselßies (Zygoptera), Ichneumoni-dae, C. sexmaculatus, Araneae, Scelionidae, red ants(Formicidae), Sphecidae, black crickets [Metiochevit-taticolis (Stål)], Dolichopodidae, Braconidae, andtropical Þre ants (S. geminata). In the riparian areas,there were no interactions between sampling date andcrop type for anyparameter(Table2). Inaddition, there

Table 1. Seasonal equivalent-mean count per 100 plants (and transformed means � SEs) for natural enemies and arthropod diversity(�SEs) in Bt- and non-Bt corn, 2006–2009

Taxa Bt corn Non-Bt corn Probabilitya

JuneÐSept. 2006M. discolor 49.49 (7.03 � 1.15) 35.50 (5.96 � 0.82) 0.428Araneae 4.38 (2.09 � 0.36) 4.16 (2.04 � 0.39) 0.814Hymenoptera 1.69 (1.30 � 0.23) 2.74 (1.66 � 0.38) 0.476Chrysopidae 0.71 (0.84 � 0.06) 1.52 (1.23 � 0.23) 0.343Diversity index 5.44 � 0.52 5.95 � 0.62 0.336

Jan.ÐApril 2007M. discolor 6.48 (2.55 � 1.82) 1.57 (1.25 � 0.16) 0.114Chilomenes sp. 4.36 (2.09 � 0.40) 3.36 (1.83 � 0.31) 0.354Araneae 2.90 (1.70 � 0.16) 2.96 (1.72 � 0.23) 0.953Ichneumonidae 0.96 (0.98 � 0.11) 0.72 (0.85 � 0.06) 0.157Syrphidae 0.66 (0.81 � 0.06) 1.08 (1.04 � 0.19) 0.498Hymenoptera 0.57 (0.75 � 0.05) 1.27 (1.13 � 0.24) 0.376Diversity index 7.49 � 0.81 8.92 � 0.69 0.284

JuneÐSept. 2007Araneae 8.90 (2.98 � 0.25) 8.40 (2.90 � 0.27) 0.579Hymenoptera 1.75 (1.32 � 0.23) 2.33 (1.53 � 0.17) 0.388Diversity index 3.09 � 0.56 2.92 � 0.42 0.506

JuneÐSept. 2008S. geminata 126.31 (11.24 � 2.45) 24.11 (4.91 � 0.65) 0.062M. discolor 24.31 (4.93 � 0.62) 61.36 (7.83 � 0.92) 0.005b

Araneae 9.02 (3.00 � 0.33) 3.00 (1.73 � 0.27) 0.065Chrysopidae 0.97 (0.98 � 0.17) 0.50 (0.71 � 0) 0.091Diversity index 4.54 � 0.37 3.71 � 0.49 0.106

Jan.ÐApril 2009M. discolor 41.81 (6.90 � 0.87) 65.44 (8.09 � 0.79) 0.480Diversity index 3.57 � 0.45 3.38 � 0.39 0.492

All arthropod taxa data was transformed on to the �(x � 0.5) scale prior to analysis. Transformed means (and their SEs) from therepeated-measures ANOVA were based on data for three replications for each of the six sample dates. Equivalent (back-transformed) meansprovide estimates of treatment means on the original scale. Diversity index data was not transformed prior to analysis but again means andSEs are based on the data used for the repeated-measures ANOVA, i.e. three replications for each of the six sample dates.aMain-effect F-test probability from repeated-measures ANOVA. All time � treatment interactions non-signiÞcant (P � 0.05), except for

M. discolor (JuneÐSept. 2008)Ñsee text and footnote 2.b SigniÞcant date x treatment interaction effect (P 0.049). Treatment t-tests for each sampling date showed signiÞcant differences between

treatments on only one sampling date.

277

Fig. 1. PRC plots of arthropod communityin corn Þelds over Þve cropping seasons. Visualsamplings of arthropods were conducted fromJune 2006 to April 2009. Œ(Bt corn), f (Non-Btcorn). Taxa with weights between 0.5 and�0.5 are not shown because they have littleinßuence on the PRC curves.

278

were no signiÞcant differences in either the diversityindex or in the abundance of any of the natural enemiesbetween the riparian sites associated with Bt corn andthose adjacent to non-Bt corn (Table 2).

Discussion

This is the Þrst long-term monitoring study con-ducted in the Philippines to measure secondary eco-logical effects of Bt corn on arthropod community incommercial-scale farmsandnearbyriparianareas.ThePRC analyses of the Þve cropping seasons of data didnot detect any adverse impacts on arthropod commu-nity either inside the corn crop or in adjacent riparianareas. Note that the PRC method has been advocatedto measure biomonitoring data (Van den Brink and

Ter Braak 1999; Van den Brink et al. 2003, 2009) andhas been shown to provide one of the best availablemethods for detecting such effects, including the ef-fects of insecticides on community structure (e.g.,CandolÞ et al. 2004, Whitehouse et al. 2005). Previousstudies (Dively 2005, Eizaguirre et al. 2006) have re-ported the absence of adverse long-term effects of Btcorn on nontarget arthropods. A multiyear study byHead et al. (2005) on Bt cotton showed no adverseimpact on nontarget arthropod populations. The eco-logical function of nontarget arthropods was not af-fected by Bt cotton, as reported by Naranjo (2009).Storer et al. (2008) concluded that there is no widescale adverse impact on nontarget species resultingfrom the widespread use of Bt crops. In addition, thelow variance explained by crop type that was observed

Fig. 2. PRC plots of arthropod community in riparian sites adjacent to corn Þelds over Þve cropping seasons. Sweepsamplings of arthropods were conducted from June 2006 to April 2009. Œ(Adjacent to Bt corn), f (Adjacent to Non-Bt corn).Taxa with weights between 0.5 and �0.5 are not shown because they have little inßuence on the PRC curves.

279

in the current study has been documented by otherworkers. For instance, Torres and Ruberson (2006)reported very low variance associated with Bt cottonon nontarget arthropods, which correlated also withlack of signiÞcant effect on abundance.

In a previous study on long-term monitoring of Btcorn, Schorling and Freier (2006) reported that Btcorn did not have a signiÞcant impact on arthropodcommunity. Similarly, Priestly and Brownbridge(2009) concluded from their 2-yr study that environ-mental conditions and crop-management practiceshad the most signiÞcant impact on arthropod com-munity rather than crop type. In the current study, thelarge variance associated with sampling dates wascommon to all the PRC analyses of abundance datafromcornandriparian sites. Incidentally,PRCanalysiscould also be used to determine a suitable bioindicatortaxa that can be used for future monitoring studies(Frampton 2000). Results of PRC analyses from thisstudy suggest that the coccinellid beetle M. discolormay be a useful indicator group for the Philippine cornecosystem.

This species is a generalist predator and is almostalways present throughout the growing season of cornin major corn-growing regions of the country (Al-cantara 2004). The ubiquitous presence ofM. discolorin the Þeld, coupled with its omnivorous feeding habit,makes the insect highly exposed to transgenic Cry1Abprotein. Our Tier I test in the laboratory showed thatM. discolor is not sensitive to Cry1Ab protein even atdoses 20 times higher than the dose of Cry1Ab proteinfound in Bt corn pollen (E.P.A., unpublished data). Ina similar study, another species of coccinellid beetle

was not adversely affected by ingesting high concen-trations of Cry1Ab protein (Alvarez-Alfageme andBigler 2010). The results of the current study furthersupport previous observation that long-term effects onthe abundance ofM. discolor in the Þeld are unlikely.The role of the other coccinellid predator,C. sexmacu-latus, as a signiÞcant biotic agent in the control of corninsect pests has not been documented locally. Per-haps, because of its low abundance in the Þeld, C.sexmaculatusmight be an efÞcient predator only in thepresence of M. discolor (i.e., there is a synergisticinteraction between the two species) (E.P.A., unpub-lished data). Spider identiÞcation was assigned only tothe higher taxa (Araneae) because it was demon-strated previously that the assemblage of spider spe-cies is more effective in reducing prey densities thana single spider species (Greenstone and Sunderland1999, Sunderland 1999). Spider abundance (i.e., at theSuper Family level) was not adversely affected by Btcorn (Table 1). Several herbivores observed in cornmight be preyed upon by spiders. Consistent withprevious observation (Jasinski et al. 2003), hymenop-teran parasitoids were observed in Bt corn despite theabsence of corn borer, but it is unclear what theseparasitoids were doing in Bt corn.

Riparian areas close to farming communities are notonly an important haven for biodiversity but mightalso be important sources of beneÞcial arthropods forthe control of insect pests in nearby corn Þelds. Thesenatural enemies (Fig. 2) could be useful in regulatingthe population of some insect pests present not onlyinside riparian sites but also in corn. The presence ofM. discolor in riparian sites suggests that the predator

Fig. 2. (Continued).

280

is very mobile and it might warrant further investiga-tion to determine its role as a major carrier of trans-genic Cry protein to these sites. Considering the veryclose proximity of Bt corn to the riparian sites, it mightalso be possible that other corn predators (M. vittati-colis and Araneae) carry the transgenic Cry protein tothe riparian sites.

The risk to the arthropod community in these sites,more generally, appears to be negligible especially asthe presence of Bt corn near riparian sites did not havea signiÞcant adverse inßuence on the abundance ofpredatory taxa. A sensitive method to quantify theamount of transgenic Cry1Ab protein that might beingested by arthropods through the food chain in theriparian sites is necessary to establish unequivocablythe insigniÞcant risk posed by the transport of trans-genic protein to these sites. Because riparian sites areimportant source of biodiversity, they could be used asone of the indicators of biodiversity loss. The Con-vention on Biological Diversity has listed trophic in-tegrity of other ecosystems as one of the headlineindicators for monitoring loss of biodiversity (re-viewed in Walpole et al. 2010). This study has shown

that long-term near deployment of Bt corn had nodetectable adverse impact on arthropod diversity inriparian sites.

Biodiversity protection is important for sustainabledevelopment (Brundtland 1987). It provides servicesfor agriculture (Altieri 1999) and is related to ecosys-temfunctioning(Clergueet al. 2005).TheexponentialShannon index was used to compare directly the ar-thropod diversity between Bt- and non-Bt corn; therewas no signiÞcant difference in arthropod diversitywithin every cropping season between Bt- and non-Btcorn. This conÞrms the Þndings of previous studies inother countries that have shown similar diversity in-dices between Bt corn and conventional corn (Lozziaet al. 1998, Lozzia 1999, Dively and Rose 2002).

The cost of regulation of genetically modiÞed (GM)crops in the Philippines is represented by activities incompliance and support of government regulatoryrequirements (Manalo and Ramon 2007). As part ofresponsible stewardship of transgenic technology,postcommercialization monitoring is a regulatory re-quirement for GM crop-technology developers wish-ing to renew the license to market their product. The

Table 2. Seasonal equivalent-mean count/ 10 quadrats (and transformed means � SEs) for natural enemies and arthropod diversity(�SEs) in riparian sites adjacent to Bt- and non-Bt corn, 2006–2009

Taxa Bt corn Non-Bt corn Probabilitya

JuneÐSept. 2006Zygoptera 1.80 (1.34 � 0.19) 1.15 (1.07 � 0.13) 0.329Anisoptera 1.08 (1.00 � 0.10) 1.08 (0.95 � 0.10) 0.947Ichneumonidae 0.76 (0.87 � 0.09) 0.58 (0.76 � 0.04) 0.108C. sexmaculatus 0.62 (0.78 � 0.05) 0.66 (0.81 � 0.06) 0.571Diversity index 14.69 � 1.18 11.77 � 1.36 0.302

Jan.ÐApril 2007Araneae 6.15 (2.48 � 0.20) 5.06 (2.25 � 0.25) 0.758Ichneumonidae 1.60 (1.26 � 0.15) 0.85 (0.92 � 0.11) 0.249Scelionidae 1.18 (1.09 � 0.17) 0.94 (0.97 � 0.14) 0.400Anisoptera 0.86 (0.93 � 0.07) 0.79 (0.89 � 0.08) 0.742Formicidae 1.64 (1.28 � 0.24) 0.74 (0.86 � 0.07) 0.125Diversity index 10.75 � 1.35 9.98 � 0.99 0.128

JuneÐSept. 2007Araneae 9.50 (3.08 � 0.29) 9.29 (3.05 � 0.25) 0.896Ichneumonidae 1.42 (1.20 � 0.14) 1.90 (1.38 � 0.14) 0.144Sphecidae 0.72 (0.85 � 0.11) 0.57 (0.75 � 0.05) 0.455Chilomenes sp. 1.26 (1.12 � 0.15) 1.38 (1.17 � 0.16) 0.766Formicidae 2.91 (1.71 � 0.29) 2.29 (1.51 � 0.23) 0.570M. vittaticolis 1.32 (1.29 � 0.17) 1.57 (1.25 � 0.12) 0.910Dolichopodidae 0.80 (0.89 � 0.09) 0.89 (0.78 � 0.05) 0.373Diversity index 11.54 � 0.88 12.58 � 0.64 0.211

JuneÐSept. 2008Araneae 6.48 (2.54 � 0.24) 4.90 (2.21 � 0.19) 0.258M. discolor 5.37 (2.32 � 0.30) 10.42 (3.23 � 0.77) 0.475Ichneumonidae 3.07 (1.75 � 0.22) 2.36 (1.54 � 0.19) 0.503Braconidae 1.93 (1.39 � 0.22) 1.43 (1.20 � 0.17) 0.206Diversity index 16.22 � 0.71 14.58 � 1.03 0.058

Jan.ÐApril 2009M. discolor 19.13 (4.37 � 0.91) 9.44 (3.07 � 0.59) 0.354S. geminata 4.54 (2.13 � 0.32) 2.71 (1.65 � 0.23) 0.464Ichneumonidae 3.71 (1.92 � 0.20) 3.05 (1.75 � 0.17) 0.376Braconidae 2.32 (1.52 � 0.25) 1.67 (1.29 � 0.15) 0.126Sphecidae 1.85 (1.36 � 0.24) 1.22 (1.11 � 0.16) 0.137Diversity index 13.38 � 0.84 12.53 � 0.78 0.572

All arthropod taxa data was transformed on to the �(x � 0.5) scale prior to analysis. Transformed means (and their SEs) from therepeated-measures ANOVA were based on data for three replications for each of the six sample dates. Equivalent (back-transformed) meansprovide estimates of treatment means on the original scale. Diversity index data was not transformed prior to analysis but again means andSEs are based on the data used for the repeated-measures ANOVA (i.e. three replications for each of the six sample dates).aMain-effect F-test probability from repeated-measures ANOVA. All time � treatment interactions nonsigniÞcant (P � 0.05).

281

current study will be a useful reference for govern-ment regulators in the Philippines in making recom-mendations for future postcommercialization moni-toring of GM crops that are already in the regulatorypipeline and in making policy guidelines for monitor-ing schemes as a type of measure for coexistence withorganic farming (Jones 2009). The issue of coexistenceof GM crops and organic farming is currently beingactively studied in the Philippines to accommodatefreedom of choice and to aid food security.

Acknowledgments

I thank Drs B. L. Caoili, P. J. Javier, and D. J. Mostoles fortechnical assistance, and John Rogers, Research Connectionsand Consulting, Brisbane, Australia, for assistance with thepreparation of this manuscript. Funding for this study wasprovided by the Program for Biosafety Systems-Biotechnol-ogy Biodiversity Interphase-United States Aid for Interna-tional Development.

References Cited

Alcantara, E. P. 2004. Monitoring insect abundance and di-versity in Bt corn. In Impact assessment of Bt corn in thePhilippines, 2004, pp. 1Ð93. Terminal Report. Interna-tional Service for Acquisition of Agribiotech Applica-tions.

Altieri, M. A. 1999. The ecological role of biodiversity inagroecosystems. Agric. Ecosyst. Environ. 74: 19Ð31.

Alvarez-Alfageme, F., and F. Bigler. 2010. Laboratory tox-icity studies demonstrate no adverse effects of Cry1Aband Cry3Bb1 to larvae of Adalia bipunctata (Coleoptera:Coccinellidae): the importance of study design. Trans-gen. Res. 20: 467Ð479.

Brookes, G., and P. Barfoot. 2008. Global impact of biotechcrops: socio-economic and environmental effects, 1996Ð2006. AgBioForum 11: 21Ð38.

Brundtland, G. 1987. Our common future: world commis-sion on environment and development. Oxford Univer-sity Press, Oxford, United Kingdom.

Candolfi, M. P., K. Brown, C. Grimm, B. Reber, and H.Schmidli. 2004. A faunistic approach to assess potentialside-effects of genetically modiÞed Bt-corn on non-targetarthropods under Þeld conditions. Biocontrol Sci. Tech-nol. 14: 129Ð170.

Clergue, B., B. Amiaud, F. Pervanchon, F. Lasserre-Joulin,and S. Plantureax. 2005. Bioversity: function and assess-ment in agricultural areas. A review. Agron. Sust. Dev. 25:1Ð15.

Dively, G. P. 2005. Impact of transgenic VIP3A x Cry1Ablepidopteran-resistant Þeld corn on the nontarget arthro-pod community. Environ. Entomol. 34: 1267Ð1291.

Dively,G. P., andR.Rose. 2002. Effects of Bt transgenic andconventional insecticide control on non-target inverte-brate community in sweet corn, pp. 265Ð274. In Proceed-ings, First International Symposium of Biological Controlof Arthropods, 14Ð18 January 2002, Honolulu, Hawaii.U.S. Dep. Agric. Forest Service Publication FHTET-2003-05. U.S. Dep. Agric., Morgantown, WV.

Eizaguirre, M., R. Albajes, C. Lopez, J. Eras, B. Lumbierres,and X. Pons. 2006. Six years after the commercial intro-duction of Bt maize in Spain: Þeld evaluation, impact andfuture prospects. Transgen. Res. 15: 1Ð12.

Frampton,G.K. 2000. Large-scale monitoring of non-targetpesticide effects on farmland arthropods in England: thecompromise between replication and realism of scale, pp.

54Ð67. In J. J. Johnston (ed.), Pesticides and wildlife, ACSSymposium Series No. 771. The American Chemical So-ciety, Washington, DC.

Fraulo, A. B., R.McSorley, andO. E. Liburd. 2008. Effect ofthe biological control agent Neoseiulus californicus (Ac-ari: Phytoseiidae) on arthropod community structure innorth Florida strawberry Þelds. Fla. Entomol. 91: 436Ð445.

Greenstone, M. H., and K. D. Sunderland. 1999. Why asymposium on spiders in agroecosystems now? J. Arach-nol. 27: 267Ð269.

Griffiths, N. A., J. L. Tank, T. V. Royer, E. J. Rosi-Marshall,M. R. Whiles, C. P. Chambers, T. C. Frauendorf, andM.A.Evans-White. 2009. Rapid decomposition of maizedetritus in agricultural headwater streams. Ecol. Appl. 19:133Ð142.

Head,G.,W.Moar,M. Eubanks, B. Freeman, J. Ruberson, A.Hagerty, and S. Turnipseed. 2005. A multiyear, large-scale comparison of arthropod populations on commer-cially managed Bt and non-Bt cotton Þelds. Environ.Entomol. 34: 1257Ð1266.

James, C. 2011. Global status of commercialized BIO-TECH/GM crops. ISAAA Brief No. 43. ISAAA, Ithaca, NY.

Jasinski, J. R., J. B. Eisley, C. E. Young, J. Kovach, and H.Wilson. 2003. Select nontarget arthropod abundance intransgenic and nontransgenic Þeld crops in Ohio. Envi-ron. Entomol. 32: 407Ð413.

Jensen, P. D., G. P. Dively, C. M. Swan, and W. O. Lamp.2010. Exposure and nontarget effects of transgenic Btcorn debris in streams. Environ. Entomol. 39: 707Ð714.

Jones, E. 2009. Consultation on proposals for managing thecoexistence of GM, conventional and organic crops inWales. Department for Rural Affairs and Heritage, WelshAssembly Government, Aberystwyth, Wales.

Jongman,R.H.G., C.J.F. TerBraak, andO.F.R. vanTongeren(eds.). 1995. Data analysis in community and landscapeecology. Cambridge University Press, Cambridge, UnitedKingdom.

Jost, L. 2006. Entropy and diversity. Oikos 113: 363Ð375.Lozzia, G. C. 1999. Biodiversity and structure of ground

beetle assemblages (Coleoptera, Carabidae) in Bt cornand its effects on non target insects. Boll. Zool. Agr.Bachic. Ser. II. 31: 37Ð58.

Lozzia, G. C., C. Furlanis, B.Manachini, and I. E. Rigamonti.1998. Effects of Bt-corn on Rhopalosiphum padi L.(Rhynchota Aphididae) and on its predator Chrysoperlacarnea Stephen (Neuroptera, Chysopidae). Boll. Zool.Agr. Bachic. Ser. II. 30: 153Ð164.

Manalo, A. J., and G. P. Ramon. 2007. The cost of productdevelopment of Bt corn event MON810 in the Philip-pines. AgBioForum 10: 19Ð32.

Naranjo, S. E. 2009. Impacts of Bt crops on non-target in-vertebrates and insecticide use patterns. CAB Rev. 4:1Ð23.

Perry, J. N., C.J.F. Ter Braak, P. M. Dixon, J. J. Duan, R. S.Hails, A. Huesken, M. Lavielle, M. Marvier, M. Scardi, K.Schmidt, B. Tothmeresz, F. Schaarschmidt, and H. vander Voet. 2009. Statistical aspects of environmental riskassessment of GM plants for effects on non-target organ-isms. Environ. Biosafety Res. 8: 65Ð78.

Priestly, A. L., and M. Brownbridge. 2009. Field trials toevaluate effects of Bt-transgenic silage corn expressingthe Cry1Ab insecticidal toxin on non-target soil arthro-pods in Northern New England, USA. Transgen. Res. 18:425Ð443.

Raney, T. 2006. Economic impact of transgenic crops indeveloping countries. Curr. Opin. Biotechnol. 17: 1Ð5.

282

Romeis, J., R. L.Hellmich,M. P. Candolfi, K. Carstens, A.DeSchijver, A.M.R. Gatehouse, R. A.Herman, J. E.Huesing,M. A. McLean, A. Raybould, A. M. Shelton, and A. Wag-goner. 2010. Recommendations for the design of labo-ratory studies on non-target arthropods for risk assess-ment of genetically engineered plants. Transgen. Res. 20:1Ð22.

Rosi-Marshall, E. J., J. L. Tank, T. V. Royer, M. R.Whiles, M.Evans-White,C.Chambers,N.P. Storer,G.P.Dively, andR. A. Herman. 2008. Landscape effects of insect-resis-tant genetically modiÞed crops. In J. Romeis, A. M. Shel-ton, and G. G. Kennedy (eds.), Integration of insect-resistant genetically modiÞed crops within IPMprograms. Prog. Biol. Control 5: 273Ð302.

Schorling, M., and B. Freier. 2006. Six-year monitoring ofnon-target arthropods in Bt maize (Cry1Ab) in the Eu-ropean corn borer (Ostrinia nubilalis) infestation areaOderbruch (Germany). J Verbrauch. Lebensm. 1 (Suppl.1): 106Ð108.

Sunderland, K. D. 1999. Mechanisms underlying the effectsof spiders on pest populations. J. Arachnol. 27: 308Ð316.

Tank, J. L., E. J. Rosi-Marshall, T. V. Royer, M. R. Whiles,N. A. Griffiths, T. C. Frauendorf, and D. J. Treering.2010. Occurrence of maize detritus and a transgenic in-secticidal protein (Cry1Ab) within the stream network ofan agricultural landscape. Proc. Natl. Acad. Sci. U.S.A.107: 17645Ð17650.

Ter Braak, C.J.F., and P. Smilauer. 2002. CANOCO refer-ence manual and CanoDraw for Windows userÕs guide:software for canonical community ordination, version 4.5.Microcomputer Power, Ithaca, New York.

Tian, J. C., Z. C. Liu, M. Chen, Y. Chen, X. Chen, Y. F. Peng,C.Hu, andG.Ye. 2010. Laboratory and Þeld assessmentsof prey-mediated effects of transgenic rice onUmmeliatainsecticeps (Araneida: Lyniphiidae). Environ. Entomol.39: 1369Ð1377.

Torres, J. B., and J. R. Ruberson. 2006. Abundance and di-versity of ground-dwelling arthropods of pest manage-ment importance in commercial Bt and non-Bt cottonÞelds. Ann. Appl. Biol. 150: 27Ð39.

Van den Brink, P. J., and C.J.F. Ter Braak. 1999. Principalresponse curves: analysis of time-dependent multivariateresponses of biological community to stress. Environ.Toxicol. Chem. 18: 138Ð148.

VandenBrink, P.,N.VandenBrink, andC.TerBraak. 2003.Multivariate analysis of ecotoxicological data using ordi-nation: Demonstrations of utility on the basis of variousexamples. Australas. J. Ecotoxicol. 9: 141Ð156.

VandenBrink, P. J., P. J. denBesten, A. bij deVaate, andC. J.Ter Braak. 2009. Principal response curves techniquefor the analysis of multivariate biomonitoring time series.Environ. Monit. Assess. 152: 271Ð281.

Volkmar,C.,C.Freir,A. Juen,M.Traugott, andM.Schorling.2004. Results of a 4-year plant survey and pitfall trappingin Bt maize and conventional maize Þelds regarding theoccurrence of selected arthropod taxa. Schweiz. Ento-mol. Gesell. Mitt. 27: 78Ð84.

Walpole, M., R.E.A. Almond, C. Besancon, S.H.M. Butchart,D. Campbell-Lendrum, G. M. Carr, L. Collette, N. C.Davidson, E. Dulloo, A. M. Fazel, et al. 2010. Trackingprogress toward the 2010 biodiversity target and beyond.Science 325: 1503Ð1504.

Whitehouse, M.E.A., L. J. Wilson, and G. P. Fitt. 2005. Acomparison of arthropod communities in transgenic Btand conventional cotton in Australia. Environ. Entomol.34: 1224Ð1241.

Yorobe, J. M., and C. B. Quicoy. 2006. Economic impact ofBt corn in the Philippines. Philipp. Agric. Sci. 89: 258Ð267.

Received 24 November 2010; accepted 13 July 2012.

283