Review

Six years after the commercial introduction of Bt maize

in Spain: field evaluation, impact and future prospects

Matilde Eizaguirre1, Ramon Albajes1,*, Carmen Lopez1, Jordi Eras2, Belen Lumbierres1 &

Xavier Pons11Universitat de Lleida, Centre UdL-IRTA, Rovira Roure 191, 25198, Lleida, Spain2Departament de Quımica, Universitat de Lleida, Rovira Roure 191, 25198, Lleida, Spain

Received 3 October 2005; accepted 7 October 2005

Key words: Bt maize, GMO, non-target, Sesamia, transgenic

Abstract

We carried out a 6-year-field evaluation to assess potential hazards of growing Compa�, a transgenic Btmaize variety based on the transformation event CG 00256-176. Two categories of hazards were investi-gated: the potential of the target corn borer Sesamia nonagrioides to evolve resistance to Bt maize andeffects on non-target organisms. In order to address the first hazard, dispersal capacity of the corn borerwas measured and our results indicated that larvae move to plants other than those onto which the femaleoviposited – even to plants in adjacent rows – in remarkable numbers and they do so mostly at a matureage, suggesting that mixing Bt and non-Bt seeds in the same field would not be a very useful deploymentstrategy to delay/prevent resistance. In addition, adults move among fields to mate and males may do so forup to 400 m. Three different aspects of potential non-target effects were investigated: sub-lethal effects onthe target S. nonagrioides, effects on non-target maize pests, and effects on maize-dwelling predators.Larvae collected in Bt fields at later growth stages, in which event 176 Bt maize expresses Bt toxin at sub-lethal concentrations, had longer diapause and post-diapause development than larvae collected in non-Btfields, a feature that might lead to a certain isolation between populations in both type of fields andaccelerate Bt resistance evolution. Transgenic maize did not have a negative impact on non-target pestsin the field; more aphids and leafhoppers but similar numbers of cutworms and wireworms were countedin Bt versus non-Bt fields; in any case differences in damage or yield were recorded. We observed nodifference in the numbers of the most relevant predators in fields containing transgenic or no transgenicmaize.

Introduction

Cultivation of transgenic crops has increasedsteadily since the commercial introduction of thefirst genetically modified crops in 1996 reaching81 million hectares in 2004 (James, 2004). Adop-tion rate of the technology is even higher if one

considers that transgenic crops are concentratedmainly in the Americas (over 90% of the globaltransgenic surface area). China and India arepoised to become leading countries for cultivationof transgenic crops in the next few years. Twotraits, resistance to insect pests and herbicidetolerance represent over 95% of all transgenictraits. In contrast, European Union regulationshad until very recently imposed a strict morato-rium on the deployment of any transgenic crops,*Author for correspondence

E-mail: ramon.albajes@irta.es

Transgenic Research (2006) 15:1–12 � Springer 2006DOI 10.1007/s11248-005-3998-1

with Spain being the only notable exception.Transgenic maize cultivation in Spain reached58,000 ha in 2004, all devoted to insect resis-tant maize engineered with genes from Bacillusthuringiensis Berliner, Bt maize.

Bt maize was first approved for commercialcultivation in Spain in 1998. Since then and until2002 only one commercial variety was grown,Compa CB� (Event 176, Cry1Ab toxin, Syngen-ta Seeds), a variety developed from the hybridDracma� (Syngenta Seeds). In the 2004 and2005 seasons, however, event MON810 has beenprevalent. Recently (August 2005) registration ofevent 176 has expired in Spain. Event 176expresses the toxin in green tissues, pollen andstalks, but not in the silks and kernels (Cannon,2000) and its concentration in the tissue progres-sively decreases after pollination until harvest(Fearing et al., 1997). Bt maize hybrids, primarilydeveloped to control the European corn borer,Ostrinia nubilalis (Hubner), have also demon-strated good control of the Mediterranean cornborer, Sesamia nonagrioides Lefebvre (author’sunpublished results). Use of Bt corn increasescrop yield to variable degrees depending on theseverity of corn borer attack; in small plots inNorth East Spain 8–14% yield increases withevents 176 and MON810 have been recordedfrom 1998 to 2004 (Lopez & Serra, personalcommunication). However, Bt maize has beenscrutinised for potential unintended effects onnon-target species, the possibility (unproven) ofantibiotic resistance markers and other geneticelements of procaryotic origin present in thetransformation construct being transferred toother organisms, transgene flow to wild relativesof maize, and potential development of resistanceto Bt in target insects.

The entomopathogenic bacterium B. thuringi-ensis has been used for over 50 years with noreported negative environmental impact. Its ento-mopathogenicity is based on the production oftoxins (insecticidal crystal proteins-Cry proteins)that are considered safe for humans and most non-pest species. Selectivity of Cry toxins is due to themechanism of activation of the insecticidal proteinin the insect gut and to the presence of specificreceptors in the midgut epithelium of target insects.Persistence of Bt toxins expressed in transgenicinsect resistant crops during the growing season isconsiderably higher compared to Bt field sprays.

The active ingredients in Bt sprays applied conven-tionally in the field are deactivated rapidly by UVand insecticidal action rarely exceeds a few days.

In the context of biosafety evaluation, risk isdefined as exposure*hazard. Four major catego-ries of potential hazard in the context of transgeniccrops have been put forward by the USA NationalResearch Council (NRC, 2002): (i) evolution ofresistance (in this case development of resistantinsect populations upon prolonged exposure to Bttoxins), (ii) movement of transgenes (with subse-quent expression in a different organism or spe-cies), (iii) whole plants (the transgenic plant itselfmay become an environmental hazard because thetraits it receives may improve its fitness), and (iv)non-target effects. As part of the environmentalrisk assessment required by Spanish governmentregulations for widespread cultivation of Bt maizeseveral R&D projects have addressed the fourcategories of hazard. Here we review those dealingwith categories (i) and (iv) with reference to thetransgenic Bt maize event 176, Compa CB�.

Concerning the first category of hazard,evolution of resistance, research has concentratedon the estimation of dispersal capacity ofS. nonagrioides in order to generate data to aidevaluation of the potential of insects to evolveresistance to Bt, and also to optimise non-Bt refugestrategies. This is well understood for O. nubilalisbut not for other corn borers. The study of thefourth hazard category, non-target effects, includeseffects of Bt exposure of S. nonagrioides to sub-lethal levels of the toxin, effects on non-target pestsin maize fields, and effects on predatory fauna.

Towards strategies to prevent evolution

of resistance to Bt in corn borers: measurement

of the dispersal capacity of S. nonagrioidesto inform decisions on refugia deployment strategies

Resistance in insects to pest control agents,particularly insecticides, is an old problem. Morethan 500 insect and mite species have been foundto be resistant to one or more insecticide activeingredients (Anonymous, 2005). Development ofresistance to Bt toxin application has been dem-onstrated in a number of insects, includingO. nubilalis, in the laboratory (Tabashnik, 1994;Chaufaux et al., 2001) although it has been doc-umented in the field only for the diamond backmoth, Plutella xylostella (L.) (Tabashnik et al.,

2

1990). The deployment of Bt maize – thatcurrently expresses one Bt toxin, Cry1Ab, contin-uously for several months – in Spain may exert ahigher pressure on susceptible herbivores withthe potential to enhance Bt resistance develop-ment in exposed corn borer populations. Sesamianonagrioides may be more likely to develop resis-tance to Bt as it is less polyphagous and moresedentary than O. nubilalis (Anglade, 1972; Eizag-uirre & Albajes, 1989). However, none of thesetwo corn borers showed consistent shifts in sus-ceptibility after 5 years of Bt maize cultivation inSpain when systematic field monitoring was car-ried out (Farinos et al., 2004). As the surface areaof transgenic Bt maize with a unique Bt toxinincreases in Spain, current recommendations andlikely future legislation in the coming years mayrequire cultivation of non-Bt maize in a fraction ofthe surface of farms growing their transgeniccounterpart to delay development of resistance toBt in corn borers. Consequently, studies such asours are invaluable in providing the regulatorswith accurate scientific information to allow themto develop and implement informed and sensiblelegislation for the prudent deployment of insectresistant transgenic crops.

Theoretical predictive models have thus farplayed a major role in the analysis of the potentialevolution of resistance in susceptible insect popu-lations and to recommend deployment strategiesto delay such resistance (Gould, 1998). As a resultof such analysis the resistance management strat-egy most commonly used in the deployment of Btcorn is the combination of a high toxin expressionin the host plant and the establishment of refugesof non-transgenic crops – a variable percentage ofthe total area planted with the crop but usuallyaround a 20%. However, assumptions underlyingmodels have been judged as excessively simplified(Vacher et al., 2003) or inaccurate, and parametersused inaccurately qualified and quantified(Gressel, 2005) and with little experimental sup-port or field validation. This last author evenquestions if actual high-dose-refuge strategy isreally needed to delay development of resistance toBt in target pests taking into account that noresistance has evolved in maize and cotton wherefarmers did not comply with the modelled recom-mendations for refuges. Availability of field datato plug into simulations may give more realisticresults for prediction of Bt resistance evolution.

When analysing evolution of resistance ordevising resistance-management strategies in cornborers it is important to account for corn boreradults and larval movement as discussed by theCanadian Expert Panel on the Future of FoodBiotechnology (Royal Society of Canada, 2001).Two key questions need to be addressed: (i) ismating random? (ii) How far do adults fly from thefield where they emerge to find a mate, or do theymove from one field to another to mate? Answersto these questions could determine how rapidlyresistance can be expected to evolve and alsorefuge usefulness, size and placement. Refugesmay consist of entire fields of non-Bt maize oralternatively fields sown with mixed Bt and non-Btseeds. In the later case, larvae eating non-trans-genic plants will not be exposed to Bt toxin andmay thus be a reservoir of susceptibility genes.This strategy, however, is only useful if larvae donot move from one plant to another during theirdevelopment or if they do move, such movement isa low probability event (Gould, 1998). All theseaspects of the biology of S. nonagrioides are poorlydocumented, consequently two types of fieldexperiments were conducted to study dispersal ofthe species. In the first set of experiments westudied larval dispersal by infesting maize plantswith borer egg masses and then we scouted theabundance and age of larvae on the original plantthat was infested and also surrounding plants, anumber of days after infestation. In the second setof experiments we used two techniques to monitoradult dispersal; one with directional light traps toestimate exchanges between adjacent fields and theother with rubidium-marked individuals andrecapture with light and pheromone traps locatedat increasing distances from the source.

Larval dispersal

We studied dispersal capacity of larvae ofS. nonagrioides in the field for 3 years (Eizaguirreet al., 2004). We selected 30 sites in one field eachyear. One site consisted of three rows of the cornfield each with 11 plants. The central plant in thecentral row was infested with a borer egg masswith a known number of eggs just before hatchingto avoid predation as much as possible; theremaining plants in each site were dissected 7, 14,and 32 days after infestation, to record the numberof young (L1–L2), intermediate age (L3–L4), and

3

mature larvae (L5–L6) or pupae. Plants of 10 siteswere thus dissected at each of the three samplingdates and year.

Larvae hatched on one plant migrated toneighbouring plants mostly at a mature age. Only1.0±0.4% of young larvae dispersed, whereas thisvalue was 8.3±0.7% for larvae of intermediate ageand reached 66.8±23.3% for mature larvae. Thesevalues, however, were demonstrated to be density-dependent and exact percentages at each develop-mental stage may thus change with populationdensity as a result of other factors related with thequality and age of the host plant. Among larvaethat moved, 43% remained in the same row and57% moved to the adjacent rows. Movement oflarvae was not in any particular direction. Thedecrease in numbers of dispersed larvae per plantas the distance to the infested plant increased wasfitted to a negative exponential function (Figure 1)that allows to estimate the average distance movedby larvae reaching the pupal stage (r=0.31 m), themedian distance (r0.50=0.22 m) and alsor0.95=0.87 m. These density–distance relationshipsshow that more than half of the larvae may moveto plants other than the plant on which theyhatched (a common between-plant distance inmaize is about 0.17 m) and they can thus beexposed to Bt toxin or that a noticeably number oflarvae may be expected to move between adjacentrows.

Movement of S. nonagrioides aged larvae to Btplants from non-Bt plants can increase the fitnessof partially resistant genotypes through enhancedsurvival of aged larvae developed on non-Bt plants

and dispersed to Bt plants. This might limit theeffectiveness of refuges based on the seed mixtureof Bt and non-Bt corn. Strip planting (alternatingBt and non-Bt rows) could also be ineffectivebecause several of the larvae that survive until lateinstars can move to rows adjacent to the ovipos-ited plant. Surrounding Bt fields with non-Bt stripsas it is frequently recommended may also favoursurvival of partially resistant individuals and thisshould be avoided.

Adult dispersal

We studied the movement of S. nonagrioides adultsfor up to 5 years in the field using two differenttechniques. The first was based on the fact that Bttransgenic maize (event 176) does not allow thedevelopment or the survival of the corn borerlarvae until early September when the first damageis observed (author’s observations). This meansthat adults in a Bt maize field during the seasonmust come from adjacent fields. Consequently,dispersal distances of the corn borer were esti-mated by monitoring movement of adults from anon-Bt field (source) to an adjacent Bt field (sink).This was carried out with directional light trapslocated in the border between the two types offields; half of the traps were oriented to Bt fields tocatch adults coming from these fields and the otherhalf were oriented to non-Bt fields to catch adultsleaving non-Bt fields (Eizaguirre et al., 2004). Withthis technique, the proportion of males remainingin the field in which they emerged compared tothose that moved to the adjacent Bt field wereestimated by comparing pheromone trap catchesin the two fields. The second technique to studymale dispersal used moths marked with rubidiumin one field (source) sprayed with RbCl. Larvae ofthe corn borer that feed in this field acquire andretain the marker until the adult stage. Rubidiumcontent in marked adults caught in pheromoneand light traps located at increasing distances fromthe source field was analysed by flame atomicemission spectrometry (Albajes et al., 2004) andnumbers of adults caught at each distance com-pared.

Catches in directional light traps were notdifferent in those oriented to Bt or non-Bt fields(Figure 2) (Eizaguirre et al., 2004) showing thatdispersal of S. nonagrioides adults was highenough to distribute them randomly between Bt

Figure 1. Number of larvae of S. nonagrioides per plant onplants located at increasing distances from the initially in-fested plant. The decrease in the number of larvae at increas-ingly distant plants is fitted by means of a negativeexponential curve, u=ae)bm, where a is the intercept on theY-axis and b the slope (Eizaguirre et al., 2004).

4

and non-Bt fields. This was also the conclusionfrom pheromone trap data for the second adultflight but not for the third. Whereas in the firstcase catches in Bt and non-Bt fields were similar,more adults were caught in traps located in thenon-Bt field in the second case. Most femalescaught in light traps were mated, irrespective oforigin (Bt or non-Bt field). This would confirm thehypothesis derived from laboratory observationsof mating behaviour in which females, unlikemales, did not move before mating and matedmostly in the first scotophase after emergence(Lopez et al., 2003). Earlier preliminary fieldobservations had already shown that femalesemerging from overwintered larvae were mostly(>94%) mated when trapped in light traps (Lopezet al., 1999). Consequently, flow of susceptibilitygenes can be attributed mostly to males whichwere studied with more detail with the Rb-markedindividuals.

In earlier work, we had concluded that pher-omone trap catches of S. nonagrioides males werenot lower at a distance of 400 m from the source ofRb-marked individuals versus the same sourcefield (Albajes et al., 2004). Two additional years ofresults using this technique confirmed our conclu-sion but only for males of the second flight,whereas males of the third flight appeared to fly atshorter distances (Figure 3, unpublished results).

We can conclude that for the second flight, theprobability for a male emerging in one Bt field tomate with a female emerging in a non-Bt fieldlocated within a radius of 400 m is the same as that

of mating with a female emerging from the sameBt field. This may be used as a criterion to locate anon-Bt field refuge to delay resistance to Bt in cornborer populations. For males of the third flight thesituation does not seem to be the same becauselocation of non-Bt refuges, even adjacent to Btfields, do not assure the mating at random.However, it has been suggested that progeny ofS. nonagrioides adults of the third flight survive inextremely low numbers as most hosts for ovipo-sition and subsequent larval development are dryor have a low nutritional quality, in addition to thelarval mortality due to the low temperatures whenlarvae are still young (Eizaguirre et al., 2002).Consequently, adults of the second flight, whichare randomly distributed within a radius of 400 m,would be more important in terms of susceptibilitygene flow between Bt fields and non-Bt refuges.

Sub-lethal effects on target organisms: the case

of S. nonagrioides

Sub-lethal effects of Bt toxins on insect perfor-mance, particularly reduced feeding and delayedlarval development have been reported by manyauthors (see references in Eizaguirre et al., 2004).Interference with larval development may alterdiapause induction and termination and thus mayaffect fitness of the targeted herbivore. Corn borersmay experience sub-lethal concentrations of Bttoxin when feeding on late growth stages oftransgenic Bt maize that is based on event 176.

Figure 2. Mean total number (±SE) of adults of S. nonagrioides per trap caught in directional light traps located between twomaize fields and oriented towards the Bt and the non-Bt field in the second and third adult flights. Each value is the mean of 3traps. (Eizaguirre et al., 2004).

5

This event exhibits a declining Bt toxin expressionfrom flowering onward (Fearing et al., 1997;Onstad & Gould, 1998) so that larvae may survivein high numbers and be exposed to sub-lethallevels of Bt late in the season. In previouslaboratory studies we had confirmed that sub-lethal concentrations of Dipel� (a mixture ofmainly two Bt toxins, Cry1Ab and Cry1Ac) causeddelayed development also in S. nonagrioides larvaewhich additionally increased the number of instarsbefore pupation, and, more importantly, criticalday length for diapause induction was increased byabout 41 min (Eizaguirre et al., 2004). To deter-mine if these effects could affect diapause and post-diapause development in overwintering larvae thatsurvive in autumn on senescent transgenic maize,these variables were studied in the field.

Effect of sub-lethal Bt on diapause and post-diapause development was measured on diapaus-ing larvae collected from a Bt transgenic (event176) and non-Bt maize (near isogenic variety)field. Those larvae were maintained at 15�C andsubjected to several day lengths and the number ofdays until pupation allowed us to compare thediapause (in larvae collected in October) andpost-diapause development [in larvae collected inFebruary when diapause had already terminated(Lopez et al., 1995, 2001)] in larvae that may havebeen exposed to sub-lethal levels of Bt.

We did not find any significant differences inlarval mortality in insects from Bt and non-Btfields. Field results were consistent with those inthe laboratory (Eizaguirre et al., 2004). In larvaecollected in October and in February, the numberof supernumerary moults before pupation washigher in larvae from the Bt field. In addition,diapause and post-diapause development periods

were found to be longer in larvae that had beenexposed to a sub-lethal concentration of Bt in thefield at most of the day lengths tested (Figure 4).This modification in the developmental and dia-pause responses to environment may decrease thefitness in those populations ingesting sub-lethalconcentrations of Bt in senescent transgenic maizeas diapause is misadjusted to real environmentalconditions. Additionally, phenology of the cornborer populations feeding on transgenic maizewith low expression of Bt toxins is altered incomparison to insects feeding on non-transgenicplants leading to a certain reproductive isolation

Figure 3. Mean number (+SE) of rubidium-marked males of S. nonagrioides caught in pheromone traps placed at increasingdistances from the plot sprayed with rubidium chloride. Each value is the average of 2 replications and 7 (in flight 2) and 8 (inflight 3) recording weeks. Bars within each block followed by the same or no letters are not significantly (p>0.05) different.

Figure 4. Duration (mean±SE days until pupation) ofdiapause (a) and post-diapause (b) development of diapausinglarvae of S. nonagrioides collected in October and February,respectively and subjected to 15�C and several different pho-toperiod regimens. Within each photoperiodic regimen, col-umns with different letters indicate different duration for (a)and (b) between non-Bt and Bt field collected larvae(Eizaguirre et al., 2005).

6

between the two populations and consequently Btresistance evolution may be more rapid.

Effects on non-target pests

Corn borers are the most harmful pest in maize inthe Mediterranean basin but other herbivores mayaffect crop yield as well, for example soil dwellinginsects (mainly wireworms and cutworms) andhomopterans. The Bt Cry1Ab toxin is very selec-tive for Lepidoptera and therefore the impact of Btmaize on non-Lepidopteran pests is expected to beminimal. However, differential expression of thetoxin in Bt maize during the life cycle of the trans-genic plant, the removal of all or part of thebiocenotic fraction that is targeted by the trans-genic crop (corn borers in this case), and potentialchanges in nutritional quality make it necessary toassess its impact fully. Direct effects of Bt-maizeon non-target pests can easily be measured inlaboratory experiments. In contrast, the dynamiceffects of the removal of other pests can only bemeasured in the field in large plots. Moreover, asCrawley (1999) stated, we need to study the effectsof genetically modified crops on the demographyof non-target species over their entire life cycle andseveral generations in the field. Such a study wascarried out at farm scale for 3 years and the majornon-target pests and predators were monitored inBt versus non-Bt plots.

Our experiments were carried out over a3-year-period in 8 adjacent commercial plots eachyear according to the crop rotation of the farm(Pons et al., 2005). The experimental design con-sisted of four randomised blocks with two treat-ments: Bt transgenic maize (Compa CB�, event176) and non-Bt near isogenic cultivar (Dracma�,Syngenta Seeds). Cultural practices applied werethose common in the area but with no soilinsecticide treatments to avoid potential interfer-ence with the trial. Because commercial plots wereused, the size varied from 0.4 to 1 ha, but wasuniform within each block. The occurrence ofpests was monitored from seedling emergence tothe onset of maturity [Hanway’s (1966) stage 9.1].

Aphids

Aphid density was evaluated by visual countingon 10–25 plants per plot. Sampling dates were

determined by the population dynamics of aphidsin the area (Asın & Pons, 2001) and they weredistributed over the entire season. For each plant,the number of aphids of each species was countedand the developmental stage was determined bydistinguishing between alate and apterous adults,alatiform and apteriform fourth instar nymphs,and young nymphs (instars 1–3).

The major findings of our experiments arereported in Pons et al. (2005). Cultivar (Bt ver-sus non-Bt) affected the total number of aphids(the four main species were Rhopalosiphum padiL., Sitobion avenae Fabricius, Metopolophiumdirhodum Walker, and Macrosiphum euphorbiaeThomas), which was significantly higher on theBt-transgenic cultivar. But when the main specieswere considered separately, only S. avenae showedsignificantly higher numbers on the Bt cultivar.The influence of the cultivar was analysed for eachdevelopmental stage of the four most abundantspecies and aphid densities were consistently high-er in Bt plots for all species and ages (except inthree out of the 25 species and age combinations).In 11 cases the differences were statistically signif-icant (Table 1). Adults (alate and apterous) andyoung nymphs of R. padi, the apterous adults andthe apteriform fourth instar nymphs of S. avenaeand adults (alate and apterous) and the apteriformfourth instar nymphs of M. dirhodum were signif-icantly more abundant on the transgenic cultivarparticularly on younger plants.

Performane of R. padi on Bt and non-Bt maizewas studied in the field and in the laboratory inmore detail (Lumbierres et al., 2004). In the field,results were in accordance with those presentedabove and Bt maize had a positive influence onaphid settlement with more adult alates and youngnymphs. As it has been well established that Bttoxin is not expressed in the phloem in event 176(Head et al., 2001; Raps et al., 2001), other factorsaffecting the process of aphid settlement or reten-tion on plants, such as host attraction or plantstructure, should be considered. Field resultsagreed with those obtained in the laboratory.Some differences in aphid performance on Bt ornon-Bt host in the laboratory were only observedin the early generations similarly to field observa-tions, where differences in aphid abundance onboth cultivars were not found from mid seasononward. The differences in aphid development onthe transgenic and near isogenic cultivars were

7

linked to changes in host-plant quality due topleiotropic effects rather than to the direct effect ofBt toxin.

Leafhoppers

Leafhoppers (Zyginidia scutellarisHerrich-Schafer)were monitored directly and indirectly twice in theseason, 10–15 days before and after anthesis. Inthe direct monitoring, three leaves were removedcarefully from the base, from three vertical strata(10 plants per plot) and brought to the lab where

the number of adults and nymphs were counted,distinguishing between young and mature nymphs.In the indirect monitoring, damage caused by theleafhoppers to the leaves was measured with aSPAD chlorophyll meter on one leaf from thesame three strata as used in the direct monitoring(but from a different set of 10 plants per plot). Atthe same time, the length and maximum width ofeach leaf were recorded and the leaf area wascalculated.

The cultivar had a significant effect on totalnumber of leafhoppers or mature nymphs. The

Table 1. Aphid densities in Bt versus non-Bt plots

Aphid species Morph Cultivar ANOVA values

Bt transgenic (n=71) Non-transgenic (n=71) F p

R. padi AL 0.57±0.19 0.37±0.14 7.97 0.02

AP 0.17±0.06 0.07±0.02 5.87 0.04

N4AL 0.22±0.22 0.27±0.24 0.02 0.89

N4AP 0.51±0.18 0.42±0.18 0.07 0.80

N1–3 2.52±0.64 1.58±0.51 5.27 0.04

Total 3.99±0.96 2.72±0.76 1.21 0.30

S. avenae AL 0.19±0.05 0.16±0.04 0.02 0.89

AP 0.31±0.09 0.14±0.04 4.02 0.08

N4AL 1.10±0.78 0.61±0.43 0.07 0.80

N4AP 0.77±0.21 0.43±0.15 6.61 0.03

N1–3 4.10±1.10 3.13±0.90 2.85 0.13

Total 6.47±1.69 4.47±1.21 2.17 0.17

M. dirhodum AL 1.18±0.36 0.78±0.24 2.63 0.14

AP 0.49±0.18 0.23±0.10 4.76 0.06

N4AL 0.72±0.26 0.67±0.25 0.01 0.96

N4AP 2.57±0.91 2.00±0.90 35.04 <0.01

N1-3 28.80±9.40 25.00±8.61 2.25 0.16

Total 33.82±10.68 28.68±9.90 3.17 0.11

M. euphorbiae AL 0.20±0.05 0.15±0.04 0.82 0.39

AP 0.39±0.13 0.33±0.12 0.05 0.83

N4AL 0.18±0.08 0.11±0.08 0.85 0.38

N4AP 0.95±0.39 1.08±0.52 0.80 0.40

N1–3 4.43±1.48 3.91±1.51 1.54 0.25

Total 6.14±2.07 5.58±2.18 1.33 0.28

Sum of 4 species AL 2.16±0.52 1.46±0.35 5.76 0.04

AP 1.36±0.41 0.77±0.24 4.82 0.05

N4AL 2.24±0.88 1.66±0.61 0.48 0.50

N4AP 4.87±1.58 3.93±1.65 3.22 0.10

N1–3 40.45±12.35 33.63±11.19 4.21 0.07

Total 51.07±14.76 41.45±13.48 5.05 0.05

Total Aphids 51.18±14.67 41.53±13.48 4.96 0.05

Mean values (±SE) of density (individuals/plant) of the different morphs (AL, alate adults; AP, apterous adults; N4AL, alatiformfourth instar nymphs; N4AP, apteriform fourth instar nymphs and N1–3, first to third instar nymphs) for the four most abundant aphidspecies (R. padi+S. avenae+M. dirhodum+M. euphorbiae) separately, and for the total aphid density of these four species in Bttransgenic and non-transgenic plots. Data were transformed to SQRT (x+0.5) before analysis (From Pons et al., 2005).

8

density in Bt transgenic plots was higher than intheir non-transgenic counterparts, but no cultivareffects were detected on the other developmentalstages (Table 2). Damage caused by leafhoppers(measured through leaf area and SPAD) was notaffected by cultivar.

As in the case of aphids, it is difficult toattribute higher densities of leafhoppers in Btmaize to the transgenic trait itself. Other investi-gators did not find significant concentrations of Bttoxin in leafhoppers feeding on Bt maize in thelaboratory (Dutton et al., 2004) or in the field(Obrist et al., 2005, in press) in spite that leafhop-pers feed on mesophyll cells of maize where thetoxin is expressed. In fact, Rauschen et al. (2004)did not find a significant difference betweenleafhopper abundance on transgenic Bt-maize(event MON810 that expresses higher toxin con-centrations) and its isogenic variety. Differentialleafhoppers densities do not appear to be directlylinked to the Bt toxin and may be irrelevant forleafhopper pest status as there were no differencesbetween the Bt and non-Bt plots in terms of thedamage caused by the herbivore.

Soil insect pests

To estimate the percentage of plants killed by soilinsects (wireworms and cutworms), the number ofplants emerged and attacked in two central rows ineach plot were recorded weekly until plants were atthe 11-leaf stage, after which no further attack wasobserved.

No other species than the cutworm A. segetumand the wireworm A. lineatus were recorded. No

effects of cultivar on the percentage of plants killedby soil worms were found.

Effects on predators

Potential effects of Bt maize on predators that areknown to suppress herbivore pest populations oncrops have been one of the main lines of investi-gation in conjunction with transgenic crops, par-ticularly after Hilbeck et al. (1998) recorded in thelaboratory some indications of negative effects offeeding lacewings on prey killed by a Bt toxin. Inthe field, predators may come into contact withtoxins produced in transgenic plants throughseveral pathways; by feeding on plant parts orpollen, through feeding on target or non-targetherbivorous insects that have ingested the toxin, orvia the environment, i.e. the soil when toxinspersist and do not lose their toxicity after plantparts or insects have died (Groot & Dicke, 2002).We have shown recently, in field studies, that Bttoxin is acquired by some herbivores that feed ontransgenic maize and by some of the predators thatfeed on prey or plant material (pollen) containingthe toxin (Obrist et al., 2005, in press). In addition,field impacts of Bt maize on predators may be theconsequence of altering the density of prey (targetand non-target herbivores as well). Some of thesepotential effects had been measured in the field,particularly in the USA but mostly in small plots(Orr & Landis, 1997; Pilcher et al., 1997). Largerplots are needed to avoid frequent populationexchanges in highly mobile predators in order togenerate meaningful data (Prasifka et al., 2005).

The same fields investigated for non-targetherbivores were used to assess effects on predators(Poza et al., 2005). Epigeal and above groundpredators were studied by both visual and pitfalltrap sampling. A variable number of 10–25 plantsper plot were inspected visually on each samplingdate (5 sample dates per season) and the numberof predators was recorded. Three pitfall traps wereinstalled in each plot and they were operative inthe field for 1 week. Pitfall trapping was carriedout for five sampling dates distributed over themaize growing season each year.

More than 90% of predators recorded in visualsamplings were Heteroptera (mostly Anthocori-dae), Coleoptera (mostly Coccinellidae), or Arach-nidae (mostly Araneae). Differences between

Table 2. Leafhopper densities in Bt versus non-Bt plots

Bt-transgenic Non-transgenic

Total leafhoppers 13.46±2.31 a 11.59±2.22 b

Adults 1.74±0.27 1.54±0.27

Mature nymphs 9.89±1.81 a 8.23±1.64 b

Young nymphs 1.82±0.36 1.82±0.49

SPAD 46.24±1.44 46.21±1.55

Leaf area 403.57±19.03 393.21±18.66

Mean values±SE of the density (individuals/plant) of theleafhopper Z. scutellaris [(total, adults, mature nymphs (N3–N5instars) and young nymphs (N1–N2 instars)] and mean±SE ofthe SPAD measures and leaf area (cm2) of Bt transgenic andnon-transgenic plants. Means are compared within each row.Values followed by different letters are significantly different(p<0.05) (From Pons et al., 2005).

9

transgenic and non-transgenic plots were notconsistent during the 3 years of our studies(Figure 5). Similarly, no significant differenceswere found in pitfall trap catches that containedmostly Carabidae, Araneae, and Dermaptera.Influence of Bt maize on epigeal predator abun-dance could have been potentially caused by oneor more of the following: the positive effect of theincreased number of prey (aphids and leafhoppers)observed in Bt plots, or the negative effects due to

the ingestion by Anthocoridae of Bt-toxin-express-ing pollen or prey with Bt-toxin (Obrist et al.,2005, in press). Exclusion of corn borer in Bt plotscould additionally deprive some of the predatorsof a part of its prey. We concluded that none ofthese potential mechanisms affected predatorabundance in the field or the different mechanismscould compensate each other in their oppositeeffects. For above ground predators, there were nodifferences in pitfall trap catches despite contactwith Bt toxin in exudates or remaining fromprevious Bt maize debris, or ingestion with preyexposed to Bt (Saxena et al., 2002). The experi-ment was replicated in Central Spain with similarresults (Poza et al., 2005). Lack of negative effectsof Bt maize on epigeal or above ground predatorsis in agreement with experiments carried out in thelaboratory with O. majusculus where nymphaldevelopment and survival was not affected whenthe predator diet consisted of Ephestiaeggs supplemented with either transgenic or non-transgenic plants and pollen (Pons et al., 2004).We use this generalist predator as a model insectbecause it may be exposed to the toxin throughmultiple and simultaneous pathways as it feeds onseveral prey species and on plant sap and pollenand it is present in maize fields throughout theseason. Our field results also agree with thosereported in field experiments conducted with Btmaize and impact on predators in the USA (e.g.Orr & Landis, 1997; Pilcher et al., 1997, 2005;Musser & Shelton, 2003; Dale & Buntin, 2005;Dively, 2005; Lopez et al., 2005).

The results of field experiments conducted atthe farm scale in Spain over a number of yearsindicate that Bt maize is compatible with theconservation of natural enemies that are respon-sible for keeping many herbivores of maize at lownumbers.

General conclusions

Applications of biotechnology to agriculture util-ising transgenic crops have been the object ofintense debate in the last several years. Europeangovernments in particular have imposed draconianrestrictions on the release of such plants into theenvironment. In Spain, the only country in theE.U. with a significant area of transgenic cropcultivation, the central government has sponsored

Figure 5. Mean (±SE) number of individual predators perplant, or per trap and week, in Bt and non-Bt plots recordedin visual samplings or caught in pitfall traps. Only the mostabundant predatory groups are represented. Within each year,asterisks denote significant (p<0.05) differences between thetwo treatments (Poza et al., 2005).

10

an environmental risk assessment project since1999. The ongoing project aims to provide Spanishregulatory bodies with methods and data forassessing potential hazards associated with thedeployment of Bt maize. Although some pointsstill need further research, several conclusions maybe drawn, in particular with reference to (i)movement of larvae and adults of the targetedcorn borer S. nonagrioides and sub-lethal effects inrelation to development of Bt resistance and (ii)effects on non-target pests and predators.

Resistance to Bt in the target corn borers inSpain, has not developed as of 2003, after 5 yearsof growing Bt maize, but recent substantialincreases in its cultivation has compelled thegovernment to impose the compulsory use ofrefugia at farm scale. In this respect, our resultsdemonstrate that separate fields may be moreadvisable than mixing Bt and non-Bt seed. Refugefields may be located at a maximal distance of400 m from transgenic ones to assure a randomdistribution of S. nonagrioides males. The useful-ness of refugia, however, may not be as substantialas currently believed. This is because survivingadults that have been exposed to sub-lethal con-centrations of the toxin in transgenic maize mayhave an asynchronous development compared toindividuals originating from non-transgenic plantsand this would limit the random mating betweenthe two populations. In addition, individualsexposed to sub-lethal levels of Bt may be largelyunfit due to the alteration of their response toenvironmental cues for diapause induction andtermination.

Lack of negative impact of Bt maize (Compa)on predators and non-target pests is in completeagreement with practically all research conductedin the field and reported in the literature. In ourexperiments, we strived to reduce the probabilityof migration of individuals between experimentaladjacent plots by using larger plots, closer in sizeto commercial fields, a fact that sets our studyapart from most previously reported studies. Inaddition, experiments were also replicated in thecentre of Spain with the same protocol we used inour location (Catalonia, NE of Iberian Peninsula)and results were the same. It is impossible toreproduce experimentally all the complexity ofrelations, mechanisms and species that potentiallymay influence the impact of Bt maize on non-tar-get fauna but the scenario used here – comparison

with a near isogenic variety, in the field, undercommercial growing conditions and over a 3-year-period – has allowed us to perform our analysisunder a realistic and relevant scenario.

Acknowledgements

This work was partially funded by the SpanishMinistry of Education and Science, Projects Nos.AGF99-0782 and AGL2002-00204.

References

Albajes R, Eras J, Lopez C, Ferran X, Vigata J and EizaguirreM (2004) Testing rubidium marking for measuring adultdispersal of the corn borer Sesamia nonagrioides: first results.IOBC/WPRS Bull 27(3): 15–22.

Anglade P (1972) Les Sesamia. In: Balachowsky AS (eds),Entomologie appliquee a l’agriculture, Tome II Lepidopteres.(pp. 1389–1401) Masson et Cie Press, Paris.

Anonymous (2005) Resistant Pest Management: ArthropodDatabase. http://www.pesticideresistance.org/DB/species.-php. Consulted on June 23, 2005.

Asın L and Pons X (2001) Effect of high temperature on thegrowth and reproduction of corn aphids (Homoptera: Aph-ididae) and implications for their population dynamics on thenortheastern Iberian Peninsula. Environ Entomol 30: 1127–1134.

Cannon RJC (2000) Bt transgenic crops: risks and benefits. IntPest Manage Rev 5: 151–173.

Chaufaux J, Seguin M, Swanson JJ, Bourget D and SiegfriedBD (2001) Chronic exposure of the European corn borer(Lepidoptera: Crambidae) to CryIAb Bacillus thuringiensistoxin. J Econ Entomol 94: 1564–1570.

Crawley MJ (1999) Bollworms, genes, and ecologists. Nature400: 501–502.

Dale T and Buntin GD (2005) Effects of Bacillus thuringiensiscorn for Lepidopteran control on nontarget arthropods.Environ Entomol 34: 1292–1301.

Dively GP (2005) Impact of transgenic VIP3A�Cry 1AbLepidopteran-resistant field corn on nontarget arthropodcommunity. Environ Entomol 34: 1267–1291.

Dutton A, Obrist L, D’Alessandro M, Diener L, Muller M,Romeis J and Bigler F (2004) Tracking Bt-toxin in transgenicmaize to assess the risks on non-target arthropods. IOBC/WPRS Bull 27(3): 57–64.

Eizaguirre M, Lopez C and Sans A (2002) Maize phenologyinfluences field diapause induction of Sesamia nonagrioides(Lepidoptera: Noctuidae). Bull Entomol Res 92: 439–443.

Eizaguirre M, Lopez C and Albajes R (2004) Dispersal capacityin the Mediterranean corn borer, Sesamia nonagrioides(Lepidoptera: Noctuidae). Entomol Exp Appl 113: 25–34.

Eizaguirre M, Lopez C and Albajes R (2005) Sublethal effectsof Bacillusthuringiensis on larval development in Sesamianonagrioides. J Econ Entomol 98: 237–247.

Eizaguirre R and Albajes R (1989) Present situation of arthro-pod pests in maize in Northeast Spain. Acta PhytopatholEntomol Hung 24: 77–80.

Farinos GP, De la Poza M, Hernandez-Crespo P, Ortego Fand Castanera P (2004) Resistance monitoring of field

11

populations of the corn borers Sesamia nonagrioides andOstrinia nubilalis after 5 years of Bt maize cultivation inSpain. Entomol Exp Appl 110: 23–30.

Fearing PL, Brown D, Vlachos D, Meghji M and Privalle L(1997) Quantitative analysis of CryIA(b) expression in Btmaize plants, tissues, and silage and stability of expressionover successive generations. Mol Breeding 3: 169–176.

Gould F (1998) Sustainability of transgenic insecticidal culti-vars: integrating pest genetics and ecology. Annu Rev Ento-mol 43: 701–726.

Gressel J (2005) Problems in qualifying and quantifying assump-tions in plant protection models: resultant simulations can bemistaken by a factor of million. Crop Prot 24: 1007–1015.

Groot AT and Dicke M (2002) Insect-resistant transgenicplants in a multi-trophic context. Plant J 31: 387–406.

Hanway JJ (1966) How a Corn Plant Develops. Special ReportNo. 48 Iowa State University of Science and TechnologyCooperative Extension Services, Ames, Iowa.

Head G, Brown CR, Groth M and Duan JJ (2001) Cry1Abprotein levels in phytophagous insects feeding on transgeniccorn: implications for secondary exposure risk assessment.Entomol Exp Appl 99: 37–45.

HilbeckA, BaumgartnerM, FriedMF and Bigler F (1998) Effectsof transgenic Bacillus thuringiensis corn-fed prey on mortalityand development time of immature Chrysoperla carnea (Nue-roptera: Chrysopidae). Environ Entomol 27: 480–487.

James C (2004) Global Status of Commercialized Biotech/GMCrops: 2003. ISAAA Briefs No. 32 International Service forthe Acquisition of Agri-Biotech Applications, Ithaca, NY.

Lopez C, Eizaguirre M and Albajes R (1995) Diapause detec-tion and monitoring in the Mediterranean corn borer. PhysiolEntomol 20: 330–336.

LopezC,EizaguirreMandAlbajesR (2003)Courtship andmatingbehaviour of the Mediterranean corn borer, Sesamia nonagrio-ides (Lepidoptera: Noctuidae). Span J Agric Res 1: 43–51.

LopezC, SansAandEizaguirreM (1999) Influencia de la plantademaız en el apareamiento de Sesamia nonagrioides Lefebvre(Lepidoptera:Noctuidae). InvestAgrProdProtVeg14: 415–422.

Lopez C, Sans A, Asın L and Eizaguirre M (2001) Phenologicalmodel for Sesamia nonagrioides (Lepidoptera: Noctuidae).Environ Entomol 30: 23–30.

Lopez MD, Prasifka JR, Bruck DJ and Lewis LC (2005) Utilityof ground beetle species in field tests of potential nontargeteffects of Bt crops. Environ Entomol 34: 1317–1324.

Lumbierres B, Albajes R and Pons X (2004) Transgenic Btmaize and Rhopalosiphum padi (Hom., Aphididae) perfor-mance. Ecol Entomol 29: 309–317.

NRC (National Research Council) (2002) Environmentaleffects of transgenic plants: the scope and adequacy of reg-ulation National Academy Press, Washington.

Musser FR and Shelton AM (2003) Bt sweet corn and selectiveinsecticides: impacts on pests and predators. J Econ Entomol96: 71–80.

Obrist L, Dutton A, Albajes R and Bigler F (2005) Exposure ofarthropod predators to Cry1Ab toxin in Bt maize fields. EcolEntomol (in press).

Onstad DW and Gould F (1998) Modelling the dynamics ofadaptation to transgenic maize by European corn borer(Lepidoptera: Pyralidae). J Econ Entomol 91: 585–593.

Orr DB and Landis DA (1997) Oviposition of European cornborer (Lepidoptera: Pyralidae) and impact of natural enemypopulations in transgenic versus isogenic corn. J EconEntomol 90: 905–909.

Pilcher CD, Obrycki JJ, Rice ME and Lewis LC (1997) Pre-imaginal development, survival and field abundance of insectpredators on transgenic Bacillus thuringiensis corn. EnvironEntomol 26: 446–454.

Pilcher CD, Rice ME and Obrycki JJ (2005) Impact of trans-genic Bacillus thuringiensis corn and crop phenology on fivenon target arthropods. Environ Entomol 34: 1302–1316.

Pons X, Lumbierres B, Lopez C and Albajes R (2004) No ef-fects of Bt maize on Orius majusculus (Heteroptera: Antho-coridae). IOBC/WPRS Bull 27(3): 131–136.

Pons X, Lumbierres B, Lopez C and Albajes R (2005) Abun-dance of non-target pests in transgenic Bt-maize: a farm scalestudy. Eur J Entomol 102: 73–79.

de la Poza M, Pons X, Farinos GP, Lopez C, Ortego F, Eiz-aguirre M, Castanera P and Albajes R (2005) Impact of farm-scale Bt maize on abundance of predatory arthropods inSpain. Crop Prot 24: 677–684.

Prasifka JR, Hellmich RL, Dively GP and Lewis LC (2005)Assessing the effects of pest management on nontargetarthropods: the influence of plot size and isolation. EnvironEntomol 34: 1181–1192.

Raps A, Kehr J, Gugerli P, Moar WJ, Bigler F and Hilbeck A(2001) Immunological analysis of phloem sap of Bacillusthuringiensis corn and of the non target herbivore Rhopalo-siphum padi (Homoptera: Aphididae) for the presence ofCry1Ab. Mol Ecol 10: 525–533.

Rauschen S, Ecker J, Gathmann A and Schphan I (2004)Impact of growing Bt-maize on cicadas: diversity, abundanceand methods. IOBC/WPRS Bull 27(3): 137–142.

Royal Society of Canada (2001) Elements of precaution: rec-ommendations for the regulation of food biotechnology inCanada. The Royal Society of Canada, Ottawa. http://www.rsc.ca//files/publications/expert_panels/foodbiotechnol-ogy/GMreportEN.pdf Consulted on June 22, 2005, page 160and others.

Saxena D, Flores S and Stotzky G (2002) Bt toxin is releasedin root exudates from 12 transgenic corn hybrids repre-senting three transformation events. Soil Biol Biochem 34:133–137.

Tabashnik BE (1994) Evolution of resistance to Bacillus thur-ingiensis. Annu Rev Entomol 39: 47–79.

Tabashnik BE, Cushing NL, Finson N and Johnson MW(1990) Field development of resistance to Bacillus thuringi-ensis in diamondback moth (Lepidoptera: Plutellidae).J Econ Entomol 83: 1671–1676.

Vacher C, Bourguet D, Rousset F, Chevillon C and HochbergME (2003) Modelling the spatial configuration of refuges fora sustainable control of pests: a case study of Bt cotton.J Evolution Biol 16: 378–387.

12