Risk assessment of Cry toxins of Bacillus thuringiensison the predatory mites Euseius concordis and Neoseiuluscalifornicus (Acari: Phytoseiidae)

Thiago Rodrigues de Castro • John Jairo Saldarriaga Ausique •

Daiane Heloisa Nunes • Fernando Henrique Ibanhes • Italo Delalibera Junior

Received: 26 April 2012 / Accepted: 18 September 2012 / Published online: 2 October 2012� Springer Science+Business Media Dordrecht 2012

Abstract Genetically modified plants carrying Cry toxins of Bacillus thuringiensis (Bt)

are widely used for pest control. Possible adverse effects as a result of the use of this

control technique to non-target organisms is still a concern; however, few studies have

addressed the effects of Bt crops on phytoseiid predatory mites. Phytoseiids are important

for the natural control of phytophagous mites, but they can also feed on pollen, plant

exudates, etc. Thus, phytoseiids may ingest Bt toxins through several pathways. In this

paper, we evaluate the direct effect of Bt-toxins by feeding the predators on Bt cell

suspensions, on solution of a Bt toxin and the tri-trophic effect by Bt expressed in trans-

genic plants. We present a method of conducting toxicological tests with Phytoseiidae

which can be useful in studies of risk analysis of toxins to be expressed by genetically

engineered plants. This method was used to evaluate the potential effect of ingestion of

suspensions of Bt (1.25 9 108 spores/ml) and of purified protein Cry1Ia12 (0.006 mg/ml

and 0.018 mg/ml) on Euseius concordis, a predatory mite that develops and reproduces

best on pollen. The effects of genetically modified Bollgard� cotton, which carries the

Cry1Ac protein, on Neoseiulus californicus, a selective predator that feeds more on spider

mites than on pollen or insects, was determined by feeding them with Tetranychus urticaereared in Bollgard� cotton and on the non-transgenic isoline. When E. concordis was fed

with suspension of Bt isolate derived from product Dipel� PM, no significant effects were

detected. Similarly, Cry1Ia12 Bt toxin, at a concentration of 0.006 mg/ml, did not affect

E. concordis. At a concentration of 0.018 mg/ml, however, the intake of this protein

reduced the reproduction of E. concordis. There were no effects of Bollgard� cotton on the

biological traits and on the predatory capacity of N. californicus. Results indicate that the

Cry toxins of B. thuringiensis studied, at the concentrations used in the field or expressed in

transgenic plants, should not affect the predatory mites E. concordis and N. californicus.

Keywords Tritrophic interaction � Transgenic cotton � Risk assessment � Biosecurity

T. R. de Castro � J. J. S. Ausique � D. H. Nunes � F. H. Ibanhes � I. Delalibera Junior (&)Departamento de Entomologia e Acarologia, Escola Superior de Agricultura ‘Luiz de Queiroz’(ESALQ), University of Sao Paulo (USP), Piracicaba, SP 13418-900, Brazile-mail: delalibera@usp.br

123

Exp Appl Acarol (2013) 59:421–433DOI 10.1007/s10493-012-9620-3

Introduction

Bacillus thuringiensis based biopesticides have been widely used to control agricultural

pests since the early 60 s (Schnepf et al. 1998). In Brazil, 26 pests of forests, crops and

vegetables are controlled with Bt based products (Polanczyk et al. 2008). Since 1996,

genetically modified plants carrying cry genes conferring resistance to pests, are used on

increasing areas worldwide, resulting in reduction of insecticide applications (Shelton et al.

2002; Wilson et al. 2004; Wu and Guo 2003). The mode of action of biological insecticides

differs from that of transgenic plants, while bio-pesticides contain the proteins in the form

of inactive pro-toxins, requiring proper pH and specific proteases of the midgut of the hosts

for protoxins cleavage and formation of toxic subunits; whereas the transgenic plants have

genes that encode Bt active toxins in soluble form (in favorable conditions of susceptible

insects, e. g. alkaline pH). These toxins bind to receptors on the peritrophic membrane and

form pores that allow the entry of Bt spores into the hemolymph, resulting in the death of

the host (Stotzky 2000).

The application of Bt bio-pesticides and genetically modified plants containing Cry

toxins can cause adverse effects on non-target organisms, including natural enemies

(Chapman and Hoy 1991; Baur and Boethel 2003; Ponsard et al. 2002; Hilbeck et al. 1998;

Dutton et al. 2003).

In Brazil, the first variety to be released for commercial planting was the Bollgard�

cotton, and more recently, a variety of cotton Bollgard II has been approved (CTNBio

2005), which presents resistance to certain species of Lepidopterans. A variety of trans-

genic cotton is being developed at EMBRAPA Genetic Resources and Biotechnology, by

inserting, in addition to the Cry1Ac protein for the control of caterpillars A. argillacea,

P. gossypiella, H. virescens (Perlak et al. 1990), the Cry1Ia12 Bt toxin, which also confers

resistance to the boll weevil (Anthonomus grandis) (Magalhaes 2006).

The potential effect of B. thuringiensis toxin on natural enemies may occur directly or

indirectly. For example, omnivorous predators may be exposed to plant toxins in different

ways: when fed on prey containing these toxins or feeding directly on plant parts, such as

pollen and/or extrafloral nectar exudates. Two species of phytoseiidae mites, Neoseiuluscalifornicus (McGregor) an specialist mite and Euseius concordis (Chant) a generalist

mite, which are found in various crops including cotton and that present very distinct

feeding habits, were selected for the assessment on safety to non-target organisms with the

use of Bt. These phytoseiids stand out for their importance in biological control in a

number of crops. N. californicus has been used in biological control programs of phy-

tophagous mites on ornamental plants, vegetables, crops in greenhouse and field crops of

citrus, corn, cassava, grape and strawberry (McMurtry et al. 1978; McMurtry and Croft

1997). N. californicus, a predator mite specialist on Tetranychidae (McMurtry and Croft

1997), can intake the Cry toxin by ingesting prey that feed on Bt plants (Dutton et al. 2002;

Obrist et al. 2006b, c). Euseius concordis as a generalist predator (McMurtry and Croft

1997) can acquire the Cry toxins, mainly, by direct consumption in the plant (pollen, nectar

and extrafloral exudates), but also indirectly by feeding on other mites and insects (trito-

phic exposure).

Despite the increasing use of genetically modified plants resistant to insects, only a few

studies have been conducted to assess the possible effects of this technology on mites,

including decomposers, phytophagous and predators (Yu et al. 1997; Oliveira et al. 2007;

Carter et al. 2004; Rovenska et al. 2005; Obrist et al. 2006a; Lozzia et al. 2000). Two

studies were carried out on the effects of the Cry toxin on the biological traits and food

preference of the predatory mites Neoseiulus cucumeris (Oudemans) and Phytoseiulus

422 Exp Appl Acarol (2013) 59:421–433

123

persimilis Athias-Henriot, respectively (Obrist et al. 2006a; Rovenska et al. 2005). Pred-

atory mites should receive high priority during the selection process of non-target species

to be evaluated in risk assessments of Bt plants due to (a) their various pathways of toxins

intake, (b) the high concentrations of toxins that they may be exposed to and (c) their

important role in biological control.

Studies have shown that levels of the Cry toxin observed in spider mites are up to 16.8

times higher than those found in tissues of transgenic plants on which they feed (Torres and

Ruberson 2008). Alvarez-Alfageme et al. 2011 found no adverse effect in the predator

Adalia bipunctata fed with spider mites (T. urticae) reared in Bt maize expressing Cry1Ab

and Cry3Bb1 fact confirmed by the ingestion of a sucrose solution containing dissolved

purified proteins at concentrations 10 times higher than measured in Bt maize fed spider

mites. The predator ladybird beetle Stethorus punctillum was not harmed when fed with

T. urticae reared in Bt maize expressing Cry3Bb1 showing detrimental effects on this

predator on Bt maize fields are unlikely (Li and Romeis 2010). As found for the other

predators, no adverse effect was found for the predatory spider Theridion impressum fed

with Cry3Bb1 containing food (pray or maize pollen) (Meissle and Romeis 2009). All this

studies shown a dilution at higher trophic levels for Cry toxins true the food chain.

The aim of this study was to evaluate the effects of ingestion of the Cry toxins on

E. concordis and N. californicus. Therefore, we developed a methodology for conducting

toxicological tests by ingestion. The direct effects of intake of the Cry toxins were

determined by ingestion of B. thuringiensis suspension (variety kurstaki, strain HD-1)

containing the toxins Cry1Ab, Cry1A, Cry2A and Cry1Ac and of the toxin Cry1Ia12 of

B. thuringiensis on the biological traits of E. concordis. We also evaluated whether

the protein Cry1Ac, found in the commercial variety of the genetically modified cotton

Bollgard�, indirectly affects the development, reproduction and predatory capacity of

N. californicus when fed on T. urticae reared on this variety of GM cotton.

Materials and methods

Toxicity tests with Euseius concordis

Collection and rearing of Euseius concordis

To conduct toxicity tests with the B. thuringiensis suspension and protein Cry1Ia12, we

collected adults of E. concordis from orange leaves (Citrus sinensis), on the campus of

Escola Superior de Agricultura ‘‘Luiz de Queiroz’’ (ESALQ—University of Sao Paulo) in

Piracicaba. The colony of the predatory mites was established according to the method-

ology proposed by Moraes and McMurtry (1981).

Each rearing unit consisted of predatory mites placed in a plate of semi-flexible vinyl

PVC resin (10 9 15 cm), type ‘‘Paviflex�’’ of gray color, overlaid in polyethylene foam

(5 9 20 9 3 cm), maintained in a plastic box (26.2 9 17.7 9 8.5 cm). The edge of the

plate was covered by a strip of cotton cloth in contact with the foam, which was moistened

on a day-to-day basis with distilled water to prevent mites from escaping. Pollen was

provided daily as food substrate to mites. Pollen of castor bean (Ricinus communis L.) was

used to create the rearing stock of mites for the test with the protein Cry1Ia12 and cattail

pollen (Typha angustifolia L.) for the assay with suspension of B. thuringiensis, according

to availability. Colonies were kept in incubator at 25 �C, humidity and 12 h of

photoperiod.

Exp Appl Acarol (2013) 59:421–433 423

123

Toxicity test of suspension of Bacillus thuringiensis

Eggs from the rearing stock were transferred to individual arenas (one egg per arena),

made of Petri dishes. Hot glue (silicone) was used to divide the Petri dishes into five

compartments, each compartment separated by a groove in which the substance to be

tested was placed. Two treatments were established: (1) suspension of spore ? crystal of

B. thuringiensis, supplied at a concentration of 1.25 9 108 spores/ml placed with artificial

blue dye commonly used in foods (ColarinMix�) at a ratio of 0.015 parts of the dye to one

part of distilled water. We placed, on average, 1.4 ml of this solution into the grooves of

the arenas, and (2) control, in which only the solution with dye at same concentration

described above was offered to the mites. The B. thuringiensis used was isolated from the

formulated product Dipel� PM (with B. thuringiensis basis, kurstaki variety, HD-1 strain,

formulated by ‘‘Hokko of Brazil’’—Municipality of Salto de Pirapora—SP, in which

containing toxins Cry1Ab, Cry1A, Cry2A and Cry1Ac). To isolate it, the product was

heated to 80 �C for 10 min to eliminate other microorganisms, non-sporulating bacteria

and to carry out the plating of dilutions in series in the culture medium for B. thuringiensis[1.0 g of glucose, 8.0 g of nutritive broth, 0.02 g of FeSO4, 0.02 g of ZnSO4, 2.0 g of

yeast, 0.02 g of MnSO4, 0.3 g of MnSO4, 12.0 g of agar and penicillin (40 mg/L)], with

pH adjusted to 7.5 (Valicente and Barreto 2003). The plates were incubated at 30 �C for

24 h. A sample of each colony was taken and observed under a microscope of phase

contrast, to confirm that only pure colonies of Bt were isolated. The plates with the

sporulated bacteria were chosen to prepare the suspension test. The culture medium was

scraped and the bacteria removed after the addition of distilled water to prepare the

concentrated suspension. The concentration of the suspension was determined with the aid

of a Neubauer chamber, being subsequently diluted to the concentration to be tested

(1.25 9 108 spores/ml). All grooves of the same Petri dish were filled by a single treat-

ment. In each arena, we placed a piece of cattail pollen and a female of E. concordis. After

confirming the oviposition, the females were removed from the arenas. The dye was used

to prove that the mites ingested the solution, by observing the blue color of the digestive

tract after ingestion, visualized through the transparency of the body of the mite (Fig. 1a).

The solution of each arena was completed along all assessments with distilled water until

the initial volume was reached (estimated by the level of blue dye left in the grooves of the

dish). After hatching, the Petri dishes containing the larvae of predatory mites remained

outside the incubator under relative humidity of about 40 % for 24 h. To maintain water

balance, the mites ingested the solution to be tested and in this case, they ingested the

B. thuringiensis suspension. Those which did not ingest it were discarded from the

experiment. After the period of low humidity, the Petri dishes were transferred to plastic

trays (6 9 22 9 14.5 cm) containing foam moistened with distilled water and with a

screen in the central part of the lid, to maintain high humidity and gas exchange inside the

plate. The trays were acclimatized in incubator at a temperature of 25 ± 1.6 �C, relative

humidity of 80 ± 16 % and 12 h of photoperiod. Measurements were made every 12 h

during the immature stages. The duration of egg stage was not considered in this exper-

iment because the female was not exposed to the treatment prior initial oviposition. After

adult emergence, we performed the assessments every 24 h. The females were mated with

males from the rearing stock. Males were caged separately. In the adult stage, we assessed

longevity of females and males and daily oviposition. A portion of cattail pollen was

placed with predatory mites on a daily basis throughout the study (to immature and adult

mites). The experiment was carried out in completely randomized design with two treat-

ments (suspension of spore ? crystal of B. thuringiensis and control) and 30 replicates.

424 Exp Appl Acarol (2013) 59:421–433

123

Toxicity test of Cry1Ia12

The methodology used was similar to that in the previous experiments. Two treatments

were established, each with its respective control group (without toxin): (1) Cry1Ia12 toxin

at a concentration of 0.006 mg/ml, and (2) Cry1Ia12 toxin at a concentration of 0.018 mg/ml.

The heterologous production method and the purification of the protein Cry1Ia12 are

described in Grossi-de-Sa et al. (2007). The toxin Cry1Ia12 was offered along with arti-

ficial blue dye and distilled water. In the control, we placed only 1–1.8 ml of blue dye and

distilled water, at a ratio of 10 lg of dye to 1 ml of water, into grooves of the arena (Petri

dish). A portion of castor bean pollen was supplied to predatory mites in both treatments,

on a daily basis (immature and adult mites). Assessments were carried out in the same way

as mentioned in the previous experiments. The experimental units were kept at a tem-

perature of 25 ± 2 �C, relative humidity of 65 ± 5 % and 12 h of photoperiod. The

experiment was performed in three different times in a completely randomized design,

totaling 38 replicates for a concentration of 0.006 mg/ml and 38 for its control group, 68

replicates for a concentration of 0.018 mg/ml and 65 for its control.

Comparison of biological features of Neoseiulus californicus in Bt and non-Bt cotton

Cultivation of cotton plants

The seeds of Bollgard� cotton (Delta Pine 4049) expressing the Cry1Ac gene and the non-

Bt isoline (Delta Pine 404) were sown in black plastic pots (15 l) inside screenhouses.

Fertilization was performed at sowing of seeds and soil coverage, as recommended for the

procedures to grow cotton. We carried out the thinning of plants, leaving two seedlings per

pot. Plants were watered twice a day, throughout their development. Plants with

Fig. 1 a Arena used to conduct toxicological tests with Euseius concordis. The suspension test was offeredalong with artificial coloring (Colormix�) shown in two compartments, while in the other compartment onlywater was offered (control) b Euseius concordis after ingestion of Bacillus thuringiensis suspensionshowing the blue digestive tract. (Color figure online)

Exp Appl Acarol (2013) 59:421–433 425

123

approximately 100 days were used to conduct the tests of the biological features and

predatory capacity of N. californicus.

Rearing method of Tetranychus urticae

Tetranychus urticae was obtained from Canavalia ensiformis L. leaves, kept in the Aca-

rology Laboratory at ESALQ-USP. Each rearing unit of T. urticae was composed of a plate

type ‘‘Paviflex�’’ placed on polyethylene foam and kept in plastic boxes. The edges of the

petioles of cotton leaves were wrapped in cotton cloth and placed on the Paviflex�plate,

maintaining the edge with cotton over the moistened foam. Afterwards, adult females of

T. urticae were transferred to start building the stock in each variety of cotton. Two

populations of T. urticae were maintained, being one permanent in Bollgard� cotton and

the other in the isoline. Every 3 days, new cotton leaves were placed over the old leaves.

The rearing of T. urticae was maintained at room temperature and a photoperiod of 12 h

for several consecutive generations, on the respective cotton varieties.

Rearing method of Neoseiulus californicus

The predatory mite N. californicus was obtained at ESALQ-USP. The units of mass rearing

and the methodology were the same used for the rearing of T. urticae. However, for the

rearing of N. californicus, cotton leaves infested with T. urticae were placed on the

‘‘Paviflex�’’ plate. Fifty adult females of the predatory mites were transferred to begin

building stock in each variety of cotton. Two populations of N. californicus were kept, being

one fed on all stages of T. urticae who were reared on Bollgard� cotton and the other fed on

T. urticae kept in the non-Bt isoline. Every 3 days, new leaves of cotton plants infested with

all stages of phytophagous mites were placed over the old leaves to feed N. californicus. The

rearing was kept at room temperature (25 ± 3 �C) at a photoperiod of 12 h.

Life table of Neoseiulus californicus

About 100 adult females of N. californicus were kept for 8 h to lay eggs on leaves of Bt

cotton or on the non-Bt isoline, in a container similar to that used in the mass rearing, at

25 ± 2 �C and relative humidity of 60 ± 5 %. The eggs were then transferred to exper-

imental units, allocating one egg to each experimental unit. The experimental units con-

sisted of an acrylic arena (2.6 cm of diameter and 1.0 cm high) with a lid containing an

opening in the center enclosed by nylon mesh for ventilation. The bottom of the arena was

lined with a small amount of moistened cotton cloth, on which the leaf disc was supported,

forming a small meniscus of water between the leaf disc and the acrylic wall, to prevent

predator mites from escaping. Daily, immature stages of T. urticae (deutonymphs) were

offered to the predators, ensuring that there were always 15 phytophagous mites per arena,

never allowing food shortage. Initially, the arenas were observed every 12 h to determine

survival and duration of each immature stage (egg, larvae, protonymph and deutonymphs).

After adult emergence, the assessments were performed every 24 h. Males were put with

females to determine longevity and fecundity. The females were mated with males from

the rearing colony when it was not enough born males, and observed on a daily basis to

determine fecundity and longevity. Mated males that died were replaced by others from the

rearing colony. To determine the sex ratio of F1 generation, all eggs were grouped daily in

an arena of the same type as the ones previously used and the mites were reared until adult

426 Exp Appl Acarol (2013) 59:421–433

123

emergence. The arenas were maintained at 25 ± 2 �C, 60 ± 5 % RH and photoperiod of

12 h. We used a completely randomized design with 26 and 27 replicates for Bt cotton and

its isoline, respectively.

Predatory capacity of Neoseiulus californicus

The experiment was conducted on leaf discs of Bt cotton and non-Bt isoline in experi-

mental units and under environmental conditions similar to those employed in the test of

the biological features of N. californicus. The females of the predatory mites used were

taken from rearing colonies kept without standardization of age. The predatory activity of

N. californicus was evaluated in a low and high host density population. One female of the

predatory mite for ten immature phytophagous mites reared in varieties of Bt cotton and

isoline was used for the low density, and for the high population density, each predatory

mite was offered daily 30 immature phytophagous mites, reared in the abovementioned

varieties of cotton. Every 24 h, we evaluated food intake, i.e., the number of phytophagous

mites preyed upon, and the number of eggs laid by the predatory mite. The predatory mites

were transferred daily to new arenas with the same initial number of preys. The arenas

were acclimatized under the same conditions presented in the previous experiment. The

total experimental period lasted 10 days. Five replicates of each treatment were established

at three different times totaling 15 replicates per treatment.

Statistical analysis

Initially, we verified the homoscedasticity of variance for data on the developmental time

of immature and adult, oviposition and predatory capacity. For homogeneous variances, we

performed the t test and for heteroscedastic variances, the Wilcoxon test, through the use of

the software program SAS 9.1 (2002–2003). The life and fertility table (LFT) for

N. californicus was prepared according to Silveira Neto et al. (1976) considering age

intervals, specific fertility and survival probability. With data on LFT, we calculated: net

reproductive rate (Ro), innate capacity to increase population (rm), finite rate of increase

(k), mean generation time (T) and doubling time (Dt), all using the Jackknife method (Maia

and Luiz 2006), also making use of the SAS program.

Results

Toxicity test of Bacillus thuringiensis suspension in Euseius concordis

No significant differences were observed in the larval, protonymph, deutonymphs stages

and larva-adult period of E. concordis between the treatment with B. thuringiensis in the

suspension and the control (without B. thuringiensis) (Table 1). Likewise, no significant

difference was found in terms of longevity of females and males. Total oviposition was

marginally significant between treatments (P = 0.051), 15.7 and 22.8 eggs laid, respec-

tively, by females that ingested the suspension with Bt and those kept in the control

treatment. When analyzing the number of eggs laid by females per day, the daily ovipo-

sition of those that ingested B. thuringiensis was similar to females that did not ingest

B. thuringiensis (P = 0.374). The longevity of females that oviposited in the Bt treatment

Exp Appl Acarol (2013) 59:421–433 427

123

was lower (14.2 days) than that of females that laid eggs in the control (19.9 days)

(P = 0.003) (data not shown in the table).

Test of toxicity of Cry1Ia12 protein in Euseius concordis

The ingestion of the toxin Cry1Ia12 at a concentration of 0.006 mg/ml did not affect the

duration of the larval, protonymph, deutonymphs stages, the larval-adult and longevity of

females and males of E. concordis (Table 2). In the reproductive phase, there was also no

significant difference between treatments in terms of oviposition and total oviposition per

female per day. When the Cry1Ia12 toxin was provided to the mites at a concentration of

0.018 mg/ml, there were no significant changes in the duration of immature and adult

stages of E. concordis (Table 3). Only the longevity of females was marginally higher in

the control treatment compared to that in the treatment with Cry1Ia12 (P = 0.068). At this

concentration of toxin, total egg production was significantly reduced (P \ 0.001) by two-

thirds compared with the control treatment. Daily oviposition was also significant, being

Table 1 The effect of ingestion of Bacillus thuringiensis (Bt) suspension on the life stage of Euseiusconcordis (mean in days ± standard error) at 25 ± 1.6 �C, 80 ± 16 % RH and 12 h of photoperiod

Parameter Control na Bt N P

Larva 0.82 ± 0.05 30 0.92 ± 0.06 30 0.201b

Protonymph 1.19 ± 0.06 29 1.17 ± 0.05 27 0.880

Deutonymph 1.03 ± 0.04 29 1.11 ± 0.05 27 0.186

Total (larva-adult) 3.01 ± 0.07 29 3.17 ± 0.06 27 0.238

Longevity of females 17.34 ± 1.44 16 14.24 ± 1.43 11 0.154

Longevity of males 13.21 ± 1.79 13 16.16 ± 1.05 13 0.168

Total oviposition 22.75 ± 2.26 12 15.68 ± 2.57 11 0.051

Eggs/female/day 1.19 ± 0.12 12 1.01 ± 0.12 11 0.374

a n = number of individuals in each life stageb Values obtained by the t test (P \ 0.05)

Table 2 The effect of ingestion of 0.006 mg/ml of Cry1Ia12 on the life stages (mean days ± SE) andoviposition rate of the mite Euseius concordis at 25 ± 2 �C, 65 ± 5 % RH and 12 h of photoperiod

Parameter Control na Cry1Ia12 N P

Larva 1.32 ± 0.12b 38 1.43 ± 0.07 38 0.738b

Protonymph 1.28 ± 0.08 37 1.36 ± 0.08 34 0.466c

Deutonymph 1.21 ± 0.08 35 1.32 ± 0.10 31 0.393c

Total (larva-adult) 3.74 ± 0.11 34 3.98 ± 0.13 31 0.160c

Longevity of females 12.85 ± 0.80 15 14.84 ± 2.16 13 0.649b

Longevity of males 7.36 ± 1.04 17 9.83 ± 1.77 16 0.231c

Total oviposition 12.87 ± 0.99 15 14.33 ± 1.45 12 0.398c

Eggs/female/day 1.19 ± 0.08 15 1.24 ± 0.11 12 0.698c

a n = number of replicates/treatmentb Value obtained by the Wilcoxon testc Values obtained by the t test

428 Exp Appl Acarol (2013) 59:421–433

123

0.61 eggs oviposited per female per day at the highest concentration of the toxin Cry1Ia12

and 1.66 egg/female/day in the control treatment (P = 0.001).

Comparative biology of Neoseiulus californicus in Bt and non-Bt cotton

The time for development of immature stages of N. californicus fed on T. urticae reared on

Bollgard� cotton and on its non-Bt isoline was similar. There were no significant differ-

ences observed between treatments regarding the longevity of males and total and daily

oviposition of females, either. The survival rate in the immature stage was 86 % in

Bollgard� cotton and 90 % in the non-Bt isoline. The sex ratio for the variety of Bt cotton

was 0.807 and 0.792 in the non-Bt isoline. The life table parameters of N. californicus,

Ro, rm, k, T and Dt for the mites fed on prey reared in Bt cotton did not differ from those

reared in the non-Bt isoline (Table 4).

Predatory capacity of Neoseiulus californicus in Bt cotton

The predatory activity of N. californicus fed on T. urticae reared on Bt cotton was not

affected, compared to that on the non-Bt isoline. When fed with a low population density

of phytophagous mites, predatory mites consumed 8.63 ± 0.17 mites/day in Bollgard�

cotton and 8.46 ± 0.21 mites/day in the non-Bt isoline (P = 0.525, t test). When predatory

mites were fed with high T. urticae density, the average number of prey consumed was

23.8 ± 0.7 mites/day in Bollgard� cotton and 24.2 ± 0.7 mites/day in the non-Bt isoline

(P = 0.700, t test). The food conversion of N. californicus was not affected by Bt toxins,

when measured by oviposition of T. urticae females fed on transgenic and non-transgenic

plants (P = 0.899, for the low population density and P = 0.224 for high population

density). The average number of eggs of N. californicus laid after ten days when this

predator was fed with low T. urticae density was 9.3 ± 0.6 and 9.4 ± 0.9 eggs per female

and when fed with high density of phytophagous mites was 10.3 ± 0.5 and 11.3 ± 0.6

eggs per female in Bt cotton and in the non-Bt isoline, respectively.

Table 3 The effect of ingestion of 0.018 mg/ml of Cry1Ia12 on the life stages (mean days ± SE) andoviposition rate of the mite Euseius concordis at 25 ± 2 �C, 65 ± 5 % RH and 12 h of photoperiod

Parameter Control na Cry1Ia12 n P

Larva 1.48 ± 0.07b 65 1.66 ± 0.10 68 0.193b

Protonymph 1.21 ± 0.08 54 1.48 ± 0.14 39 0.148b

Deutonymph 1.10 ± 0.05 51 1.23 ± 0.12 25 0.562b

Total (larva-adult) 3.73 ± 0.12 51 3.71 ± 0.18 25 0.937c

Longevity of females 10.49 ± 0.87 26 7.69 ± 1.12 12 0.068c

Longevity of males 8.84 ± 0.78 20 7.90 ± 0.59 11 0.417c

Total oviposition 15.40 ± 1.08 25 5.33 ± 1.73 06 \0.001c

Eggs/female/day 1.66 ± 0.17 25 0.61 ± 0.09 06 0.001b

a n = number of replicates/treatmentb Value obtained by the Wilcoxon testc Values obtained by the t test

Exp Appl Acarol (2013) 59:421–433 429

123

Discussion

The tests used to assess the effects of GM plants on predatory mites of the family

Phytoseiidae in the literature were based on offering phytophagous mites that were reared

on GM plants and non-transgenic plants. In this study, we present a toxicological test to

perform a risk assessment of purified Cry toxins or of bacterial suspensions to predatory

mites even before genetic modification of plants. The methodology is simple, inexpensive

and does not require the rearing of the mite to be used as prey. Although it is not possible to

determine the dose ingested by each mite, the use of a dye in solution confirms that all

individuals ingested the testing substance during the experimental period. The toxicolog-

ical test proposed is applicable in the initial studies of risk analysis to simulate the worst-

case scenario in the laboratory (tier 1), and in case negative effects are observed, it is

advisable to conduct studies in a more realistic condition in semi-field (tier 2) and, then, in

field crops (tier 3) as suggested in most protocols for risk assessment of genetically

modified plants on non-target organisms (Andow and Zwahlen 2006; Garcia-Alonso et al.

2006; Romeis et al. 2008). Risk analysis is finished when no effects are detected in

toxicological tests in each step.

The ingestion of the B. thuringiensis cell suspension (containing four Cry toxins) did not

result in significant differences in the biological features of E. concordis, except for

marginal effect on oviposition. Similarly, the toxin Cry1Ia12 did not affect the develop-

ment and reproduction of E. concordis when they were exposed to a low protein

Table 4 Duration of the life stages (days), oviposition and parameters of life and fertility table(mean ± SE) of the predatory mite Neoseiulus californicus fed with the phytophagous mite Tetranychusurticae reared in genetically modified Bollgard� cotton and the non-Bt isoline at 25 ± 2 �C, 60 ± 5 % RHand 12 h of photoperiod

Parameter Isoline na Bollgard� n P

Egg 1.90 ± 0.09 53 1.98 ± 0.09 52 0.515b

Larva 0.80 ± 0.04 53 0.77 ± 0.06 50 0.320c

Protonymph 1.64 ± 0.12 53 1.62 ± 0.13 50 0.902b

Deutonymph 1.87 ± 0.13 42 2.13 ± 0.20 36 0.346c

Total (ovo-adult) 6.24 ± 0.23 42 6.53 ± 0.31 36 0.449b

Longevity of females 27.56 ± 1.45 16 28.87 ± 1.42 15 0.527b

Longevity of males 23.23 ± 1.49 11 21.35 ± 1.28 13 0.346b

Total oviposition 35.56 ± 2.01 16 34.2 ± 2.00 15 0.634b

Oviposition/female/day 1.87 ± 0.11 16 1.81 ± 0.16 15 0.744b

Ro (female/female) 27.21 ± 1.41 23.53 ± 1.37 0.072d

rm 0.27 ± 0.01 0.24 ± 0.02 0.177d

k (female/female/day) 1.31 ± 0.02 1.27 ± 0.02 0.172d

T (days) 12.04 ± 0.50 12.92 ± 1.02 0.446d

Dt (days) 2.52 ± 0.10 2.83 ± 0.23 0.230d

a n = number of individuals in each stage of developmentb Value obtained by the t testc Values obtained by Wilcoxon testd Means did not differ from one another, by the t test (P \ 0.05) after estimating variances by Jackknifemethod

430 Exp Appl Acarol (2013) 59:421–433

123

concentration(6 lg/ml). However, the intake of 18 lg/g (0.018 mg/ml) of the toxin

Cry1Ia12 significantly affected egg production. The concentration of 18 lg/ml is higher

than the levels of other toxins found in genetically modified plants used so far and 3.5

times higher than the concentration of the toxin Cry1Ab found in T. urticae (5.13 lg/g

fresh weight) by Obrist et al. (2006a) and 7 times higher than that found in the same

phytophagous mite (2.5 lg of Cry1Ab per gram of fresh weight) by Dutton et al. (2002).

Hilbeck et al. (1998) observed high mortality rate of larvae of Chrysoperla carnea when

fed with high dose of Cry1Ac (100 lg/mL of diet), but they did not find effects on this

insect larvae when eggs of Ephestia kuehniella (Zeller) was offered concomitantly with the

same diet, demonstrating that although negative effects occur when the insect is exposed to

the purified toxin, adverse effects may not manifest when more than one source of food is

offered. Although pollen of castor bean was offered as food for E. concordis, it is possible

that under field conditions where the mites have a greater choice of food, a possible

ingestion of doses of the toxin Cry1Ia12 may not result in significant effects. E. concordisis a polyphagous predator that has multiple food sources. Species of the genus Euseius can

feed on several families of mites, including Tetranychidae, Eriophyidae and Tarsonemidae,

in addition to insects like whitefly and thrips, and pollen (McMurtry and Croft 1997).

Although E. concordis develops best on pollen, the level of expression of Cry proteins is

lower in pollen than in other parts of the plant (Kozeil et al., 1993; Greenplate, 1997) and

much lower than that found in the phytophagous mite that feeds on the transgenic plants.

The effects of the toxin Cry1Ia12 on reproduction of E. concordis indicates that additional

studies should be conducted in semi-field (tier 2) or in the field crops (tier 3), in case the

level in transgenic plants or accumulated by phytophagous mites are near the concentration

level of 0.018 mg/ml..

The results presented in the current study do not indicate any effect of genetically

modified Bollgard� cotton on the biological parameters and predatory capacity of

N. californicus, corroborating a series of studies on the safety of this technology to other

groups of non-target organisms. The use of Bt-transgenic varieties can contribute to

integrated pest management strategies and can lead to substantial reductions in insecticide

use in some crops without affecting the abundance and activity of parasitoids and predators

(Romeis et al. 2006). Meta-analyses of the extant literature on invertebrate non-target

effects carried out by Naranjo (2009) revealed that the pattern and extent of impact varies

in relation to taxonomy, ecological or anthropomorphic guild, route of exposure and the

non-Bt control against which effects are gauged. Hazards identified in the laboratory may

not always manifest in the field. Oliveira et al. (2007), for example, found that the Boll-

gard� cotton has not affected the development of decomposer mite Scheloribates prae-incisus (Berlese). Esteves Filho et al. 2010 studied the effects of Bt cotton expressing

Cry1Ac toxin (Acala DP 90B) and it’s isoline (Acala DP 90) on the compared biology and

behavior of T. urticae and Phytoseiulus macropilis (Banks) and found no effect in the

developing stages, survival, reproductive output and predatory behavior. In a study carried

out with another predatory mite, Obrist et al. (2006a) observed that there were no effects of

the toxin Cry1Ab on the biological parameters of N. cucumeris when fed with T. urticaemaintained in Bt corn (Bt11). On the other hand, Rovenska et al. (2005) showed that

P. persimilis, in a free choice test, consumed fewer T. urticae that when fed with transgenic

eggplant containing the toxin Cry3Bb than those who had consumed the non-transgenic

variety. The authors argue that this change may be due to changes in nutritional quality of

prey caused by the host plant or the recognition of the toxin in the prey, or both. It is also

possible that the effect may not be due to the presence of the Bt toxin.

Exp Appl Acarol (2013) 59:421–433 431

123

In tests conducted with two species of phytoseiidae predators and five Cry toxins found in

B. thuringiensis cells, in transgenic cotton or as purified protein, only the protein Cry1Ia12 at

the highest concentration of 0.018 mg/ml affected E. concordis. The results indicate that the

Cry toxins of B. thuringiensis studied, at the concentrations used in the field or expressed in

transgenic plants, should not affect predatory mites E. concordis and N. californicus.

Acknowledgments The first author was a recipient of a scholarship from Sao Paulo Research Foundation(FAPESP), and the study was partly funded by the Young Scientist Fellowship (FAPESP 03/00077-1)granted to Italo Delalibera Junior. We thanks Dr. Maria Fatima Grossi-de-Sa, from Embrapa RecursosGeneticos e Biotecnologia for providing the Cry1Ia12 toxin.

References

Alvarez-Alfageme F, Bigler F, Romeis J (2011) Laboratory toxicity studies demonstrate no adverse effectsof Cry1Ab and Cry3Bb1 to larvae of Adalia bipunctata (Coleoptera: Coccinellidae): the importance ofstudy design. Transgenic Res 20:467–479

Andow DA, Zwahlen C (2006) Assessing environmental risks of transgenic plants. Ecol Lett 9:196–214Baur ME, Boethel DJ (2003) Effect of Bt-cotton expressing Cry1A(c) on the survival and fecundity of two

hymenopteran parasitoids (Braconidae, Encyrtidae) in the laboratory. Biol Control 26:325–332Carter ME, Villani MG, Allee LL, Losey JE (2004) Absence of non-target effects of two Bacillus thurin-

giensis coleopteran active d-endotoxins on the bulb mite, Rhizoglypus robini (Claparede) (Acari,Acaridae). J Appl Entomol 128:56–63

Chapman MH, Hoy MA (1991) Relative toxicity of Bacillus thuringiensis var. tenebrionis to the two-spotted spider mite (Tetranychus urticae Kock) and its predator Metaseiulus occidentalis (Nesbitt)(Acari, Tetranychidae and Phytoseiidae). J Appl Entomol 111:147–154

CTNBio (Comissao Tecnica Nacional de Biosseguranca) (2005) On line consultation of technical advice.Available at http://www.ctnbio.gov.br/index.php/content/view/10951.html. Assessed in May 2009

Dutton A, Klein H, Romeis J, Bigler F (2002) Uptake of Bt-toxin by herbivores feeding on transgenic maizeand consequences for the predator Chrysoperla carnea. Ecol Entomol 27:441–447

Dutton A, Klein H, Romeis J, Bigler F (2003) Prey-mediated effects of Bacillus thuringiensis spray on thepredator Chrysoperla carnea in maize. Biol Control 26:209–215

Esteves Filho AB, Oliveira JV, Torres JB, Gondim MGC Jr (2010) Compared biology and behavior ofTetranychus urticae Koch (Acari: Tetranychidae) and Phytoseiulus macropilis (Banks) (Acari: Phy-toseiidae) on BollgardTM and non-Transgenic Isoline Cotton. Neotrop Entomol 39(3):338–344

Garcia-Alonso M, Jacobs E, Raybould A, Nickson TE, Sowig P, Willekens H, Kouwe PVD, Layton R,Amijee F, Fuentes AM, Tencalla F (2006) A tiered system for assessing the risk of geneticallymodified plants to non-target organisms. Environ Biosafety Res 5(2):57–65

Greenplate J (1997) Response to reports of early damage in 1996 commercial Bt transgenic cotton (Boll-gardTM) plantings. Soc Invertebr Pathol Newslett 29:15–18

Grossi-de-Sa MF, de Magalhaes MQ, Silva MS, Silva SMB, Dias SC, Nakasu EYT, Brunetta PSF, OliveiraGR, de Oliveira Neto OB, de Oliveira RS, Soares LHB, Ayub MAZ, Siqueira HAA, Figueira ELZ(2007) Susceptibility of Anthonomus grandis (Cotton Boll Weevil) and Spodoptera frugiperda (FallArmyworm) to a Cry1Ia-type toxin from a brazilian Bacillus thuringiensis strain. J Biochem Mol Biol40(5):773–782

Hilbeck A, Moar WJ, Pusztai-Carey M, Filippini A, Bigler F (1998) Toxicity of Bacillus thuringiensisCry1Ab toxin to the predator Chrysoperla carnea (Neuroptera: Chrysopidae). Environ Entomol27(5):1255–1263

Kozeil MG, Beland GL, Bowman C, Carozzi NB, Crenshaw R, Crossland L, Dawson J, Desai N, Hill M,Kadwell S, Launis K, Lewis K, Maddox D, McPherson K, Meghji MR, Merlin E, Rhodes R, Warren G,Wright M, Evola SV (1993) Field performance of elite transgenic maize plants expressing an insec-ticidal protein derived from Bacillus thuringiensis. Bio-Technol 11:194–200

Li Y, Romeis J (2010) Bt maize expressing Cry3Bb1 does not harm the spider mite, Tetranychus urticae, orits ladybird beetle predator, Stethorus punctillum. Biol Control 53:337–344

Lozzia GC, Rigamonti IE, Manachini B, Rocchetti R (2000) Laboratory studies on the effects of transgeniccorn on the spider mite Tetranychus urticae Koch. Boll Zool Agrar Bachic 32:35–47

Magalhaes MTQ (2006) Toxinas Cry: perspectivas para obtencao de algodao transgenico brasileiro. Dis-sertation, Universidade Ferderal do Rio Grande do Sul, RS

432 Exp Appl Acarol (2013) 59:421–433

123

Maia AHN, Luiz AJB (2006) Programa SAS para analise de tabelas de vida e fertilidade de artropodes: ometodo Jackknife. Comunicado Tecnico 33. Jaguariuna, 11 p

McMurtry JA, Croft BA (1997) Life-styles of phytoseiid mites and their roles in biological control. AnnuRev Entomol 42:291–321

McMurtry JA, Oatman ER, Phillips PA, Wood CW (1978) Establishment of Phytoseiulus persimilis (Acari:Phytoseiidae) in Southern California. Entomophaga 23(2):175–179

Meissle M, Romeis J (2009) The web-building spider Theridion impressum (Araneae: Theridiidae) is notadversely affected by Bt maize resistant to corn rootworms. Plant Biotechnol 7:645–656

Moraes GJ, McMurtry JA (1981) Biology of Amblyseius citrifolius (Denmark and Muma) (Acarina: Phy-toseiidae). Hilgardia 49(1):1–29

Naranjo SE (2009) Impacts of Bt crops on non-target invertebrates and insecticide use patterns. CAB RevPerspect Agric Vet Sci Nutr Nat Res 4(11) http://fbae.org/2009/FBAE/website/images/pdf/imporatant-publication/impacts-of-bt-crops-on-non-target-invertebrates-and-insecticide-use-patterns.pdf

Obrist LB, Klein H, Dutton A, Bigler F (2006a) Assessing the effects of Bt maize on the predatory miteNeoseiulus cucumeris. Exp Appl Acarol 38:125–139

Obrist LB, Dutton A, Albajes R, Bigler F (2006b) Exposure of arthropod predators to Cry1Ab toxin in Btmaize fields. Ecol Entomol 31:143–154

Obrist LB, Dutton A, Romeis J, Bigler F (2006c) Biological activity of Cry1Ab toxin expressed by Bt maizefollowing ingestion by herbivorous arthropods and exposure of the predator Chrysoperla carnea.Biocontrol 51:31–48

Oliveira AR, Castro TR, Capalbo DMF, Delalibera I Jr (2007) Toxicological evaluation of geneticallymodified cotton (Bollgard�) and Dipel� WP on the non-target soil mite Scheloribates praeincisus(Acari: Oribatida). Exp Appl Acarol 41:191–201

Perlak FJ, Deaton RW, Armstrong TA, Fuchs RL, Sims SR, Greenplante JT, Fischhoff DA (1990) Insectresistant cotton plants. Biotechnology 8:939–943

Polanczyk RA, Valicente FH, Barreto MR (2008) Utilizacao de Bacillus thuringiensis no controle de pragasagrıcolas na America Latina. In: Alves SB, Lopes RB (eds) Controle microbiano de pragas na AmericaLatina. Fealq, Piracicaba, pp 111–136

Ponsard S, Gutierrez AP, Mills NJ (2002) Effect of Bt-toxin (Cry1Ac) in transgenic cotton on the adultlongevity of four heteropteran predators. Environ Entomol 31(6):1197–1205

Romeis J, Meissle M, Bigler F (2006) Transgenic crops expressing Bacillus thuringiensis toxins andbiological control. Nat Biotechnol 24:63–71

Romeis J, Bartsch D, Bigler F, Candolfi MP, Gielkens MMC, Hartley SE, Hellmich RL, Huesing JE, JepsonPC, Layton R, Quemada H, Raybould A, Rose RI, Schiemann J, Sears MK, Shelton AM, Sweet J,Vaituzis Z, Wolt JD (2008) Assessment of risk of in-sect-resistant transgenic crops to nontargetarthropods. Nat Biotechnol 26:203–208

Rovenska GZ, Zemek R, Schmidt JEU, Hilbeck A (2005) Altered host plant preference of Tetranychusurticae and prey preference of its predator Phytoseiulus persimilis (Acari: Tetranychidae, Phytoseii-dae) on transgenic Cry3Bb-eggplants. Biol Control 33:293–300

SAS Institute (2002–2003) User’s manual, version 9.1.3. SAS Institute, Cary, NCSchnepf E, Crickmore N, Van Rie J, Lereclus D, Baum J, Feitelson J, Zeigler DR, Dean DH (1998) Bacillus

thuringiensis and its pesticidal crystal proteins. Microbiol Mol Biol Rev 62(3):775–806Shelton AM, Zhao JZ, Roush RT (2002) Economic, ecological, food safety, and social consequences of the

deployment of Bt transgenic plants. Annu Rev Entomol 47:845–881Silveira Neto S, Nakano O, Barbin D, Nova NAV (1976) Manual de ecologia dos insetos. Sao Paulo, CeresStotzky G (2000) Persistence and biological activity in soil of insecticidal proteins from Bacillus thurin-

giensis and of bacterial DNA bound on clays and humic acids. J Environ Qual 29:691–705Torres JB, Ruberson JR (2008) Interactions of Bacillus thuringiensis Cry1Ac toxin in genetically engineered

cotton with predatory heteropterans. Transgenic Res 17:345–354Valicente FH, Barreto MR (2003) Bacillus thuringiensis survey in Brazil: geographical distribution and

insecticidal activity against Spodoptera frugiperda (J.E. Smith) (Lepidoptera: Noctuidae). NeotropEntomol 32(4):639–644

Wilson LJ, Mensah RK, Fitt GP (2004) Implementing integrated pest management in Australian cotton. In:Horowitz AR, Ishaaya I (eds) Insect pest management: field and protect crops. Springer, Berlin, pp 97–118

Wu K, Guo Y (2003) Influences of Bacillus thuringiensis Berliner cotton planting on population dynamicsof the cotton aphid, Aphis gossypii Glover. Northern China. Environ Entomol 32(2):312–318

Yu L, Berry RE, Croft BA (1997) Effects of Bacillus thuringiensis toxins in transgenic cotton and potato onFalsomia candida (Collembola: Isotomidae) and Oppia nitens (Acari: Oribatidae). J Econ Entomol90(1):113–118

Exp Appl Acarol (2013) 59:421–433 433

123