Efects Of Bt Transgenic Cotton Lines On The Cotton Bollworm Parasitoid Microplitis Mediator In The Laboratory

Biological Control 35 (2005) 134–141 www.elsevier.com/locate/ybcon

EVects of Bt transgenic cotton lines on the cotton bollworm parasitoid Microplitis mediator in the laboratory Xiao-xia Liu a, Qing-wen Zhang a,¤, Jian-Zhou Zhao b, Jian-cheng Li c, Bao-liang Xu d, Xiao-mu Ma a a

Department of Entomology, College of Agronomy and Biotechnology, China Agricultural University, Beijing 100094, China b Department of Entomology, Cornell University-NYSAES, Geneva, NY 14456, USA c Institute of Plant Protection, Hebei Academy of Agriculture and Forestry. Baoding 071001, China d Chinese Academy of Inspection and Quarantine, Beijing 100025, China Received 2 February 2005; accepted 19 August 2005

Abstract Microplitis mediator (Haliday) is an important endoparasitoid of the cotton bollworm, Helicoverpa armigera (Hübner) in northern China. Interactions among H. armigera, its larval parasitoid M. mediator, and insect–resistant transgenic cotton lines were evaluated under laboratory conditions. Two major transgenic cotton cultivars used in Hebei province of northern China, DP99B (Bollgard), carrying the cry1Ac gene, and SGK321, carrying both cry1A and CpTI (Cowpea trypsin inhibitor) genes, were used in the experiments. The results indicated that there was signiWcant growth inhibition of the H. armigera larvae when they were fed a diet containing Bt transgenic cotton powder. The parasitoid oVspring developed more slowly and pupal and adult weight was reduced signiWcantly when the parasitized host larvae fed on the Bt cotton powder leaf diet compared with non-Bt treatment. With an increase of Bt cotton leaf powder concentration in the host larvae diet, the parasitism rate and adult emergence of the parasitoid decreased and the abnormal pupal rate increased. There was no evident diVerence in the eVects on M. mediator between the transgenic single- and two-gene cotton cultivars; however, the parasitized host larval mortality was higher than that of unparasitized larvae in most treatments. The observed eVects on M. mediator were probably host-quality mediated rather than direct eVects of transgenic cotton because the H. armigera larvae which fed on diet with leaf powder of both transgenic cotton cultivars also experienced a signiWcant decrease in weight, particularly when the host larvae were parasitized.  2005 Elsevier Inc. All rights reserved. Keywords: Microplitis mediator; Braconidae; Helicoverpa armigera: Noctuidae; Bacillus thuringiensis; Bt cotton; Transgenic plant; Parasitism rate; Non-target insect

1. Introduction The adoption of transgenic insect–resistant plants as a mean of crop protection in agriculture has proceeded rapidly since advances in recombinant DNA technology have enabled their use (James, 2004). The area of transgenic cotton expressing Cry1A insecticidal protein from Bacillus thuringiensis (Bt) increased from 2.8 million hectares in *

Corresponding author. Fax: +86 10 62734946. E-mail addresses: [email protected] (X. Liu), zhangqingwen@ 263.net (Q. Zhang). 1049-9644/$ - see front matter  2005 Elsevier Inc. All rights reserved. doi:10.1016/j.biocontrol.2005.08.006

2003 to 3.7 million hectares in China, equivalent to 66% of the total cotton producing area of 5.6 million hectares in 2004 (James, 2004). Research from both the laboratory and the Weld indicates that Bt plants are eVective in killing the larvae of a number of species of Lepidoptera, including the cotton bollworm, Helicoverpa armigera (Hübner) (Deng et al., 2003; Shelton et al., 2002; Zhao et al., 1998), a key crop pest in China that caused large production losses prior to the introduction of Bt (Wu et al., 1997). The eYcacy of Bt cotton cultivars developed both by Monsanto in the US and by Chinese institutions was excellent for the control of H. armigera (Zhao et al., 2000) and pink bollworm

X. Liu et al. / Biological Control 35 (2005) 134–141

(Pectinophora gossypiella) (Wan et al., 2004). Coincident with the increased pest control, there has been a signiWcant reduction in the amount of insecticide used, with the average number of pesticide applications per season declining from 20 to seven in Hebei and Shangdong Province (Pray et al., 2001). Theoretical models and experimental data suggest that plants expressing two dissimilar insecticidal proteins in the same plant have the potential to delay insect resistance more eVectively than single toxin plants (Zhao et al., 2003). A range of crops has been modiWed to express diVerent plant-derived genes encoding insecticidal proteins, including enzyme inhibitors, lectins, and hydrolytic enzymes (Bell et al., 2001). Cowpea trypsin inhibitor (CpTI), a plantderived gene, was expressed successfully in plant and demonstrated enhanced levels of insect resistance (Bell et al., 2001). In recent years, transgenic cotton line SGK321 expressing two insecticidal proteins (Cry1A and CpTI) (Guo et al., 1999) was commercialized in northern China (Shelton et al., 2002) and demonstrated increased insecticidal activity on H. armigera relative to single gene Bt cotton (Fan et al., 2001). Transgenic cotton plants with two Bt genes (Cry1Ac and Cry2Ab2) were approved for commercial use in Australia and the US in 2002 with superior control of cotton pests (Jackson et al., 2004). Laboratory results suggest that H. armigera is capable of developing resistance to Bt protoxins or Bt cotton over long-term selection (Akhurst et al., 2003; Liang et al., 2000; Lu et al., 2004; Meng et al., 2003; Zhao et al., 1998). At the same time, the deployment of transgenic crops must be compatible with other pest management strategies, such as biological control, if their use is to be sustainable. Additionally, it has been reported that a proportion of H. armigera larvae fed on transgenic cotton leaves in the late season can survive (Zhao and Rui, 2004), and that the surviving larvae may remain as potential hosts for parasitoid. As a result, the parasitoids could be aVected by the presence of Bt protein in food ingested by the host larvae. Accordingly, there is a need to evaluate the potential non-target eVects associated with the use of transgenic plants. A number of studies have addressed the eVect of Bt transgenic crops and Bt protoxin on beneWcial non-target insects, both parasitoids and predators. Some results showed no apparent negative eVects of transgenic plants on parasitoids (Atwood et al., 1998; Johnson, 1997; Schuler et al., 1999) some reported lower parasitoid survival rates due to premature host death (Blumberg et al., 1997), lower parasitoid emergence rates (Atwood et al., 1997), increased parasitoid larval development times (Liu et al., 2005a), or reduced longevity and fewer female ova (Baur and Boethel, 2003). However, there have been only a few reports on eVects of transgenic plants expressing two insecticidal proteins on parasitoids (Liu et al., 2005b; Ren et al., 2004). Microplitis mediator (Haliday) (Hymenoptera: Braconidae) is an important endoparasitoid of H. armigera in northern China (Liu et al., 2005a). Historically, the average parasitism rate on H. armigera was 22.9% in the cotton

135

Welds of Hebei Province (Wang et al., 1984) where insecticidal transgenic cotton cultivars have been used in almost 100% of cotton Welds in recent years. The technology necessary for the mass propagation of the parasitoid has been developed since 2002 in China, and release of the wasps to control H. armigera on non-Bt cotton Welds in Xinjiang cotton belt showed promising results (Q.-W. Zhang et al, unpublished data). The overall goal of this study was to determine the potential host-mediated eVects of Bt cotton on M. mediator under laboratory conditions. SpeciWcally, this study quantiWed the eVects of transgenic cotton expressing either one or two insecticidal proteins on selected life history parameters of M. mediator, including egg-larval period, development rate, immature stage mortality, cocoon weight, cocoon duration, adult weight and longevity. 2. Materials and methods 2.1. Insects Helicoverpa armigera was reared in the laboratory at 26 § 1 °C, 70% RH and a photoperiod of 14:10 (L:D). This species had been cultured for more than 10 years without exposure to Bt or other insecticides. The adults were fed a 10% honey solution and the larvae were reared on an artiWcial diet (Liu et al., 2004a, 2005a,b). Microplitis mediator was obtained from the Institute of Plant Protection, Hebei Academy of Agriculture and Forestry in China. The parasitoid emerges from the host as a mature larva and spins a silk cocoon inside which it pupates. Parasitized hosts do not pupate, but remain in their larval instar until the parasitoid emerges. The host subsequently dies 2–11 days after parasitoid emergence (Wang et al., 1984). The parasitoid adults fed on a 10% honey solution and mated in culture cages (20 £ 20 £ 10 cm). Females from three to eight days old were used in the experiments. The mated female wasps were allowed to parasitize H. armigera larvae, and the parasitized larvae were then reared in the laboratory on an artiWcial diet at 26 § 1 °C and 14L:10D (Liu et al., 2005a,b). 2.2. Cotton plants The transgenic cotton cultivars DP99B (Bollgard), carrying cry1Ac gene developed by Monsanto in the US, and SGK321, carrying both cry1A and CpTI (Cowpea trypsin inhibitor) gene developed by the Agri-Biotechnology Research Institute of CAAS in China (Guo et al., 1999; Zhang et al., 2004), were used in the experiments. Both cultivars are major insecticidal transgenic cotton lines used in Hebei Province in northern China. The Cry1A expression level (mean § SEM) of SGK321 (90.4 § 8.3 ng/g) was similar to that of DP99B (80.8 § 14.5 ng/g) in the cotton leaf powder used in the study ground under liquid nitrogen as tested by ELISA method (Fan et al., 2001) using the commercial ELISA QuantiPlate kit (Envirologix, USA) for

136

X. Liu et al. / Biological Control 35 (2005) 134–141

Cry1Ab/Cry1Ac quantiWcation. Shiyuan321, the parental cultivar of SGK321 and a major commercial cotton cultivar originally used in northern China, was used as the conventional cotton. The three cotton cultivars used in the experiments were provided by the Chinese Academy of Inspection and Quarantine. All plants were cultivated in greenhouses. When the plants reached 20–30 cm (7-leaf stage) in height, their leaves were taken for use in the experiments. In the preliminary bioassays using fresh transgenic cotton leaf and neonates or second instars of H. armigera, almost 100% mortality of H. armigera was observed. Larvae of M. mediator could not complete their development in host fed on fresh transgenic cotton leaf because of host death. To test for possible eVects of transgenic cotton on M. mediator, the cotton leaves were ground to powder under liquid nitrogen. The powder was stored at ¡20 °C and then added to artiWcial diet of cotton bollworm for the experiments. 2.3. Bioassays using leaf powder of transgenic cotton expressing Cry1Ac 2.3.1. EVects of Bt cotton on host larvae The Bt cotton cultivar DP99B and non-transgenic cultivars Shiyuan321 were used. H. armigera neonates were reared individually on artiWcial diets containing leaf powder of Bt or non-Bt cotton at concentrations of 0.5, 1, 3, and 5%, respectively. Each of the treated H. armigera larvae were then weighed at growth intervals of 4, 6, 8, and 10 days. Determination of the larval growth curve Wt was performed using the exponential regression function in Microsoft Excel, with time (t, day) as the dependent variable and larval weight (W, mg) as the independent variable (W D a eb, t) (Liu et al., 2005a). Larval period, pupal rate and pupal weight were all recorded, with 10 larvae serving as a replication. Six replications per treatment were performed for a total of 60 larvae per treatment. 2.3.2. EVects of Bt cotton on M. mediator Helicoverpa armigera neonates were reared individually for 4 or 5 days on the control diet containing 5% non-Bt cotton leaf powder, for 6 days on diets containing Bt cotton leaf powder at 0.5 and 1%, for 7 days at 3%, and for 8 days at 5%. Most of the treated host larvae reached the second instar and then were provided to the parasitoid. Each larva was observed until the parasitoid made one successful oviposition into the larva. Parasitized larvae were removed immediately, and were reared on the same diet they had previously. Then the next treated larva was given to M. mediator. Adult M. mediator females used in the experiment were mated and between 3 and 8 days old. The same group of females was allowed to parasitize H. armigera larvae fed on diets of diVerent treatments. At least 20 larvae from each treatment were parasitized daily, with 100 H. armigera larvae parasitized per treatment. Twenty larvae comprised a replicate and each treatment had at least Wve replicates.

The egg and larval duration period, pupal period, pupal, and fresh adult weight, and the longevity of adults (males and females) were all recorded. Adults were fed daily on 10% honey solution. Host mortality rate (HMR), successful parasitism rate (SPR), calculated based on the number of hosts still alive on the day when mature larvae of M. mediator emerged and the larvae formed a normal cocoon (Weseloh and Andreadis, 1982), abnormal cocoon rate (ACR, when the mature M. mediator larvae can emerge from host larvae, but they cannot successfully form a cocoon), and emergence rate (ER) were calculated as follows: HMR% D M/N0 £ 100 SPR% D N1/(N0 ¡ M) £ 100 ACR% D N2/(N1 + N2) £ 100 ER% D E/N1 £ 100 N0, total number of parasitized larvae; M, number of parasitized larvae that died during development; N1, number of successful cocoons; N2, number of abnormal cocoons; and E, number of emerged adults 2.4. Comparison of the eVects of transgenic cotton expressing Cry1Ac and Cry1A + CpTI To keep the size of the parasitized host larvae the same, H. armigera neonates were reared on normal artiWcial diets until they reached second instars. They were then introduced to M. mediator for parasitization according to the same procedure used in 2.3 with three cotton varieties, DP99B (single gene cotton), SGK321 (two-gene cotton) and Shiyuan321 (Control). The diet concentrations of cotton leaf powder were 2, 4, 8, and 12%, respectively. In each treatment, 80 larvae were parasitized. Half of the parasitized larvae (40) were weighed daily, mortality was calculated on the other half (40) after 8 days. A total of sixty larvae from each treatment were left unparasitized as a negative control. Thirty larvae were weighed daily after treatment, and mortality of H. armigera after 8 days was estimated from the other 30 larvae. The egg and larval duration period of the oVspring, cocoon period, cocoon weight, and adult fresh weight were recorded. 2.5. Statistical analysis One-way ANOVA was performed using SPSS (SPSS Institute, 1998). H. armigera mortality within groups (same concentration, same variety) between hosts treated by the combination of Bt-diet plus parasitism and hosts treated by Bt-diet alone and some parameters between Cry1Ac and Cry1A + CpTI cotton cultivars were compared by independent t tests at the 0.05 level. Treatment means, including abnormal cocoon rate, successful parasitism rate, adult emergence, egg-larvae duration, cocoon weight, fresh wasp weight, and wasp longevity, were compared and separated by Duncan’s Multiple Range test at P D 0.05.

X. Liu et al. / Biological Control 35 (2005) 134–141

3. Results 3.1. EVects of transgenic cotton expressing Cry1Ac 3.1.1. EVects of Bt cotton on host larvae Bt cotton leaf powder signiWcantly aVected the development of H. armigera larvae. According to the growth equation, inhibition of larval growth increased when the concentration of Bt cotton leaf powder was higher. Larval period was also signiWcantly longer compared to the control (F D 83.76, df D 191, P < 0.001), while both pupal rate (F D 3.25, df D 47, P D 0.008) and weight (F D 5.30, df D 155, P < 0.001) were signiWcant lower than the control (Table 1). No signiWcant diVerence among concentrations was observed in larval period, pupal rate and pupal weight (F D 0.15, df D 96, P D 0.927; F D 0.259, df D 23, P D 0.854; F D 77, df D 1.284, P D 0.286; respectively) when the H. armigera larvae fed on diets containing conventional cotton leaf powder. However, there were signiWcant diVerences among concentrations in larval period (F D 54.98, df D 94, P < 0.001) and pupal rate (F D 3.49, df D 23, P D 0.035) when H. armigera larvae fed on the diet containing Bt cotton leaf powder, but pupal weight was not also signiWcantly diVerent among concentrations (F D 2.330, df D 77, P D 0.081) (Table 1). 3.1.2. EVects of Bt cotton on M. mediator The eVects of Bt cotton leaf powder diets on the development of the parasitoid M. mediator are presented in Table 2. Egg-larval development was prolonged signiWcantly at all

137

concentrations compared to the control (F D 7.81, df D 314, P < 0.001). Cocoon weight in all treatments also decreased compared to the control (F D 13.05, df D 298, P < 0.001). When host larvae fed on diets containing 3% Bt cotton powder, cocoon weight decreased to 3.5 mg, while that for the control was 4.1 mg. Adult weight also decreased signiWcantly (F D 13.41, df D 243, P < 0.001). At concentrations of 1 and 3%, adult male longevity decreased signiWcantly in comparison with the control (F D 2.44, df D 171, P D 0.049), although there were no signiWcant diVerences in male and female cocoon duration times under all treatments in comparison with the control (F D 0.68, df D 172, P D 0.610 and F D 0.98, df D 71, P D 0.424, respectively). Female adult longevity was not aVected signiWcantly (F D 1.31, df D 71, P D 0.276). The host mortality varied from 12.49 to 29.82% among all treatments (Table 3). There was a signiWcant diVerence between the 3% treatment and the control (F D 8.15, df D 22, P < 0.001). When the concentration of Bt cotton leaf powder was 5%, the successful parasitism rate decreased signiWcantly compared to that observed at concentrations of 0.5% and 1%, but there were no obvious diVerences compared with the control (F D 1.95, df D 22, P D 0.145). The abnormal cocoon rate increased signiWcantly when host larvae were treated with higher concentrations of 3 and 5% (F D 7.36, df D 22, P < 0.001). The emergence rate was also signiWcantly lower under diets with 3 and 5% concentrations of Bt cotton leaf powder, but the diVerence was not signiWcant with 0.5 and 1% concentrations when compared to control (F D 2.35, df D 22, P D 0.093) (Table 3).

Table 1 Bioassay of H. armigera on diet containing non-Bt (CK) and Bt (DP 99B) cotton leaf powder Treatment (%)

0.0267t

CK-0.5 CK-1 CK-3 CK-5 Bt-0.5 Bt-1 Bt-3 Bt-5 a b

Growth equationa W D 0.3269 e W D 0.3264 e0.0264t W D 0.2426 e0.0284t W D 0.2827 e0.0272t W D 0.1925 e0.0243t W D 0.1531 e0.0229t W D 0.1047 e0.0247t W D 0.1899 e 0.0181t

Larval period (days)b

Pupal rate (%)b

Pupal weight (mg)b

15.2 § 0.4 a 14.9 § 0.4 a 15.1 § 0.4 a 15.3 § 0.4 a 17.7 § 0.4 b 19.9 § 0.5 c 22.0 § 0.4 d 25.4 § 0.5 e

71.7 § 7.0 ab 70.0 § 5.8 ab 70.0 § 7.3 ab 76.7 § 4.2 a 68.3 § 6.0 ab 66.7 § 7.6 ab 51.7 § 7.9 bc 41.7 § 5.4 c

241.0 § 10.9 a 224.4 § 7.3 ab 231.7 § 6.2 abc 220.5 § 6.5 abc 212.7 § 7.2 bc 202.0 § 7.5 cd 185.4 § 9.6 d 185.1 § 12.7 d

W, larval weight (mg); t, time (day). Means (§SEM) within the same column followed by diVerent letters are signiWcantly diVerent (P < 0.05, Duncan’s multiple range test).

Table 2 The development of M. mediator in hosts fed on diet containing non-Bt (CK) and Bt (DP99B) cotton leaf powder Treatment (%)

Egg-larval period (days)

Pupal weight (mg)

Male pupal period (days)

Female pupal period (days)

Adult weight (mg)

Male adult longevity (days)

Female adult longevity (days)

CK-5 Bt-0.5 Bt-1 Bt-3 Bt-5

7.61 § 0.05 (75) a 8.06 § 0.10 (54) b 7.89 § 0.06 (65) b 8.35 § 0.17 (62) c 7.92 § 0.07 (59) b

4.1 § 0.06 (72) a 3.8 § 0.07 (54) b 3.9 § 0.06 (64) b 3.5 § 0.06 (56) c 3.7 § 0.05 (53) b

4.37 § 0.08 (46) a 4.45 § 0.10 (24) a 4.52 § 0.09 (42) a 4.46 § 0.10 (26) a 4.54 § 0.09 (35) a

5.00 § 0.15 (20) a 4.88 § 0.12 (17) a 5.19 § 0.13 (8) a 5.16 § 0.09 (19) a 5.13 § 0.13 (8) a

1.8 § 0.04 (66) a 1.6 § 0.04 (40) b 1.6 § 0.04 (50) b 1.4 § 0.04 (45) c 1.6 § 0.03 (43) b

11.85 § 0.57 (46) a 12.00 § 0.78 (23) ac 10.10 § 0.61 (42) b 10.04 § 0.61 (26) bc 11.83 § 0.61 (35) a

17.95 § 1.60 (20) a 15.06 § 1.58 (17) a 20.75 § 2.27 (8) a 15.00 § 1.96 (19) a 18.88 § 3.18 (8) a

Means (§SEM) within the same column followed by diVerent letters are signiWcantly diVerent (P < 0.05, Duncan’s multiple range test). Numbers in brackets refer to the total number of insects used.

138

X. Liu et al. / Biological Control 35 (2005) 134–141

Table 3 EVects of diet containing non-Bt (CK) and Bt (DP 99B) cotton leaf powder on M. mediator and host larvae Treatment (%)

No. of parasitized larvae

Host larval mortality (%)

Parasitism rate (%)

Abnormal pupal rate (%)

Adult emergence (%)

CK-5 Bt-0.5 Bt-1 Bt-3 Bt-5

157 101 120 154 146

14.99 § 1.27 a 12.49 § 1.81 a 15.92 § 1.64 a 29.82 § 3.65 b 21.14 § 2.49 a

60.81 § 2.41 ab 64.24 § 1.41 a 64.97 § 6.68 a 57.85 § 3.46 ab 50.34 § 5.40 b

11.35 § 0.77 a 17.71 § 0.60 ab 13.44 § 4.79 a 34.95 § 8.41 bc 48.26 § 7.95 c

88.15 § 1.93 a 77.78 § 1.96 ab 75.90 § 6.02 ab 71.46 § 6.03 b 71.44 § 5.04 b

Means (§SEM) within the same column followed by diVerent letters are signiWcantly diVerent (P < 0.05, Duncan’s multiple range test).

3.2. Comparison of the eVects of transgenic cotton expressing Cry1Ac and Cry1A + CpTI CK DP99B SGK321

CK + P DP99B + P SGK321 + P

A 2% cotton leaf powder 100

10

B 4% cotton leaf powder 100

Mean weight per larva (mg)

3.2.1. EVects of transgenic cotton lines and parasitoid on host larvae In most of the treatments, the weight of unparasitized H. armigera larvae was more than those parasitized (Fig. 1). Seven days after treatment, the diVerence was statistically signiWcant (2%: F D 43.99, df D 69, P < 0.001; 4%: F D 49.11, df D 79, P < 0.001; 8%: F D 75.06, df D 71, P < 0.001; 12%: F D 69.70, df D 97, P < 0.001). For example, the weight of parasitized host larvae was less than 40 mg until the development of M. mediator larvae was complete, while the weight of the unparasitized larvae fed on diet containing transgenic single gene cotton (DP99B) was >100 mg. For parasitized host larvae, the diVerence of host weight was signiWcant among three cotton varieties seven days after treatment at 4, 8, and 12% concentrations (4%: F D 7.37, df D 48, P < 0.001; 8%: F D 18.50, df D 37, P < 0.001; 12%: F D 9.17, df D 57, P < 0.001). For unparasitized larvae, the diVerence of larval weight seven days after treatment was statistically signiWcant in all treated concentrations (2%: F D 17.30, df D 30, P < 0.001; 4%: F D 14.71, df D 30, P < 0.001; 8%: F D 46.56, df D 33, P < 0.001; 12%: F D 26.73, df D 39, P < 0.001). Higher growth inhibition was observed in the two-gene cotton treatment compared with the singlegene, especially when the leaf powder concentrations were higher (8 and 12%, Figs. 1C and D). The mortality of unparasitized H. armigera larvae fed on transgenic cotton diet was not signiWcantly diVerent from the control for the concentrations of 2 and 4% (2%: F D 0.16, df D 12, P D 0.853; 4%: F D 0.65, df D 12, P D 0.543), but it was signiWcantly higher than the control in transgenic cotton diet for the concentration of 8% (F D 4.396, df D 12, P D 0.043) and in the two-gene cotton treatment at 12% (F D 5.24, df D 14, P D 0.023) (Table 4). There was no signiWcantly increased mortality of parasitized host larvae in the treatments of transgenic cotton lines compared with the control at the concentrations of 2 and 4% (2%: F D 0.03, df D 13, P D 0.975; 4%: F D 0.26, df D 13, P D 0.775). Mortality of parasitized hosts was signiWcantly lower in transgenic single cotton diet than in transgenic two-gene cotton diet and control at the concentrations of 8 and 12%, but the diVerence was not signiWcant among three cotton cultivars (8%: F D 3.38, df D 13, P D 0.072; 12%: F D 3.38, df D 13, P D 0.061). For most treatments within groups (same concentration, same vari-

10

C 8% cotton leaf powder 100

10

D 12% cotton leaf powder 100

10

0

2

4

6

8

Days after treatment Fig. 1. Weight of H. armigera larvae fed on diet containing diVerent concentrations (A–D) of non-Bt (CK), Cry1Ac (DP99B) or Cry1A + CpTI (SGK321) cotton leaf powder either unparasitized or parasitized (treatment + P) by M. mediator.

X. Liu et al. / Biological Control 35 (2005) 134–141

139

Table 4 EVects of diet containing non-Bt (CK), Cry1Ac (DP99B) or Cry1A + CpTI (SGK321) cotton leaf powder on M. mediator and host larvae Concentration of cotton leaf powder (%)

2

4

8

12

Cotton varieties

CK SGK321 DP99B CK SGK321 DP99B CK SGK321 DP99B CK SGK321 DP99B

Egg-larval period (d)

7.8 § 0.1 (15) a 8.5 § 0.1 (35) b 8.6 § 0.2 (15) b 8.0 § 0.2 (12) a 8.3 § 0.1 (36) a 9.0 § 0.2 (19) b 8.3 § 0.1 (20) a 8.9 § 0.2 (28) b 8.9 § 0.2 (19) b 8.0 § 0.1 (34) a 8.6 § 0.1 (32) b 8.7 § 0.1 (27) b

Cocoon weight (mg)

4.4 § 0.2 (12) a 3.7 § 0.1 (29) b 3.8 § 0.1 (14) b 4.1 § 0.1 (11) a 3.8 § 0.1 (32) a 4.1 § 0.1 (16) a 4.2 § 0.1 (15) a 3.5 § 0.1 (27) b 4.0 § 0.1 (19) a 4.1 § 0.1 (25) a 3.6 § 0.1 (25) b 3.7 § 0.1 (26) b

Cocoon period (d)

4.7 § 0.2 (12) a 5.1 § 0.2 (23) a 5.1 § 0.3 (7) a 4.8 § 0.2 (10) a 5.1 § 0.2 (20) a 5.0 § 0.3 (12) a 4.4 § 0.2 (10) a 4.9 § 0.2 (22) a 4.7 § 0.2 (17) a 4.6 § 0.1 (19) a 4.8 § 0.2 (15) a 4.6 § 0.1 (20) a

Adult weight (mg)

Host larval mortality (%)

1.7 § 0.1 (12) ab 1.5 § 0.0 (23) a 1.7 § 0.1 (7) b 1.8 § 0.1 (9) a 1.5 § 0.1 (20) b 1.6 § 0.1 (12) ab 1.7 § 0.1 (9) a 1.5 § 0.1 (22) a 1.5 § 0.1 (17) a 1.7 § 0.0 (19) a 1.5 § 0.1 (15) b 1.5 § 0.1 (20) ab

Transgenic cotton plus parasitism 37.5 § 6.3 a 35.8 § 7.1 a 37.5 § 2.5 a 35.8 § 4.8 a 30.8 § 4.9 a 32.5 § 4.8 a 25.0 § 6.5 ab 36.7 § 4.9 b 17.5 § 4.8 a 25.2 § 2.7 a 23.3 § 6.7 a 7.5 § 4.8 b

Transgenic cotton alone 13.3 § 8.8 a 16.0 § 5.1 a 12.0 § 3.7 a** 10.0 § 5.8 a* 16.0 § 4.0 a* 16.0 § 2.5 a* 13.3 § 3.3 a 24.0 § 2.5 b 28.0 § 2.4 b 8.0 § 5.8 a* 28.0 § 3.7 b 20.0 § 3.2 ab*

Means (§SEM) within the same column and same concentration followed by diVerent letters are signiWcantly diVerent among cotton varieties (P < 0.05, Duncan multiple range test). Means within the same concentration followed by * indicates signiWcant diVerence between parasitized and unparasitized larvae (P < 0.05, independent t test). Numbers in brackets refer to the total number of insects used.

ety), parasitized larval mortality was greater than unparasitized mortality. At the 4% concentration level, the diVerence was statistically signiWcant by an independent sample test (t D 0.05 level). Parasitized host mortality was higher than that of unparasitized larvae at concentrations of 2 and 4% (t < 0.05), but signiWcantly less than that for unparasitized larvae at 12% concentration of DP 99B (t < 0.05) (Table 4). 3.2.2. EVects of transgenic cotton lines on M. mediator When the host larvae fed on diet containing leaf powder of transgenic cotton expressing Cry1Ac or Cry1A + CpTI, M. mediator oVspring had longer egg-larva duration than the control (2%: F D 6.96, df D 64, P D 0.010; 4%: F D 5.86, df D 66, P D 0.005; 8%: F D 3.58, df D 66, P D 0.034; 12%: F D 8.87, df D 92, P < 0.001) (Table 4). There was no signiWcant diVerence between the transgenic cotton lines expressing Cry1Ac and Cry1A + CpTI based on independent t test at the concentrations of 2, 8, and 12% (t D 0.05 level). At the 4% level, the egg-larval period in single gene cotton diet was longer than that in double gene cotton diet treatment (t < 0.05). For the same cotton variety, there was also no signiWcant diVerence among concentrations (CK: F D 2.38, df D 80, P D 0.076; SGK321: F D 2.16, df D 130, P D 0.096; DP 99B: F D 0.08, df D 79, P D 0.458) (Table 4). In the treatments of 2, 8, and 12% concentrations of transgenic cotton leaf powder, the cocoon weight of M. mediator was signiWcantly lower compared to the control (2%: F D 8.62, df D 54, P < 0.001; 8%: F D 18.8, df D 60, P < 0.001; 12%: F D 7.46, df D 75, P < 0.001). However, the diVerence was not statistically signiWcant at 4% (F D 2.17, df D 58, P D 0.124). The cocoon weight in 8% two-gene cotton diet treatment was signiWcantly less than that in single gene cotton diet (t < 0.05). There was no signiWcant diVerence between transgenic two-gene and single gene cotton at 2, 4, and 12% based on t tests at t D 0.05 (Table 4). For the same cotton variety, there were no signiWcant diVerences

among the various concentrations for any of the measured parameters. Cocoon period was also not aVected when the host larvae fed on diet with leaf powder of both transgenic cotton varieties in most of the treatments. The diVerence was statistically signiWcant at 2% for adult weight among cotton cultivars (F D 4.12, df D 41, P D 0.024). Adult weight in transgenic two-gene cotton was signiWcantly lower than control at 4 and 12%, but there was no signiWcant diVerence when compared to transgenic single cotton(4%: F D 2.89, df D 40, P D 0.068; 12%: F D 2.78, df D 53, P D 0.071). The diVerence was also no statistically signiWcant at 8% among three cotton cultivars (F D 2.04, df D 47, P D 0.142). 4. Discussion Helicoverpa armigera larvae are highly susceptible to intoxication from Bt transgenic cotton, making it diYcult to study the eVect of Bt cotton on the development of parasitoids if using Bt cotton leaf directly because of high mortality of host larvae fed transgenic cotton and the parasitoid larvae not being able to complete their development in dead hosts. To overcome this problem, we used ground cotton leaves added to cotton bollworm diets (Greenplate, 1999; Ren et al., 2004), then tested a range of concentrations in a preliminary experiment. Larvae of H. armigera can survive on diet containing Bt cotton leaf powder in the tested concentrations, but their larval period was longer, and pupal rate and pupal weight decreased compared with non-Bt treatment. In the present study, adult M. mediator female parasitized the host larvae fed on Bt-diet in a no-choice situation, and M. mediator larvae can complete their development in such hosts. However, immature parasitoids required more time to develop completely, both pupal and adult weight were decreased, and abnormal pupal rate increased when the concentration of Bt cotton leaf powder was higher in the diet. Further, when M. mediator developed inside host larvae fed

140

X. Liu et al. / Biological Control 35 (2005) 134–141

Bt-diets, the adult emergence rate decreased. There was no evident diVerence in the eVects on M. mediator between transgenic single- and two-gene cotton cultivars. A study with Bt cotton showed that the parasitoid Cotesia marginiventris (Cresson) developed more slowly in the soybean looper (Pseudoplusia includens Walker) fed Bt cotton than non-Bt cotton. OVspring of C. marginventris developed inside P. includens larvae fed a NuCotn 33B diet suVered reduced longevity and the females had fewer ova (Baur and Boethel, 2003). In a previous study, we used a susceptible laboratory strain and a strain collected from Bt cotton Welds as hosts to study the eVect of Bt protein on the development of M. mediator (Liu et al., 2004a). The results showed that larval period of M. mediator was delayed, and pupal weight, newly emerged adult weight and adult longevity decreased signiWcantly within the range of concentrations tested. Salama et al. (1991) reported that emergence rate and adult longevity decreased when the parasitoid Bracon brevicornis fed on host larvae treated with the diVerent concentrations of Bt. Ren et al. (2004) reported similar results that transgenic Cry1A + CpTI cotton suppressed the growth and development of host no matter whether it was parasitized or not. The pupal rate and pupal weight of M. mediator and Campoletis chlorideae Uchida (Hymenoptera: Ichneumonidae) parasitizing the host fed on transgenic cotton declined greatly. Duration of egg and larvae stage was prolonged, pupal and adult weight of C. chlorideae was decreased when H. armigera fed on transgenic Cry1A + CpTI cotton leaves compared to those fed on traditional cotton leaves for 12–48 h (Liu et al., 2005b). In the present study, there was no evident diVerence in the eVects on M. mediator between transgenic single and two-gene cotton cultivars. However, with increasing area of transgenic two-gene cultivars in China, further research, for example, host location, success with oviposition etc., is needed to study eVects on parasitoids in the future. Within the limited range of concentrations tested, H. armigera was capable of surviving on the Bt-diet, but their development was delayed signiWcantly. The delayed growth allows more time for M. mediator to parasitize the host, increasing the parasitism rate because M. mediator prefers to parasitize low instar host larvae (Wang et al., 1984). Johnson and Gould (1992) and Johnson (1997) demonstrated a synergism between transgenic plants and parasitoids, and suggested that the synergism may be the result of increased development time in Heliothis virescens fed Bttobacco. However, there is some evidence suggesting that parasitoid populations may be lower in transgenic cotton Welds compared to conventional Welds because of reduced host density (Cui and Xia, 1999). There is no general conclusion about the compatibility between Bt transgenic crop and parasitoids. Atwood et al. (1997) reported H. virescens larvae parasitized by C. marginiventris were more susceptible to Bt mixed into diet. Survival of P. includens parasitized by C. marginventris was reduced about 15% (Baur and Boethel, 2003). In our study, mortality of host larvae exposed to the

combination of Bt-diet and M. mediator parasitism was higher than that of Bt-diet only in most treatments. However, we could not conclude that parasitism by M. mediator increased the susceptibility of H. armigera to the endotoxin of transgenic cotton because of high mortality of parasitized larvae in the non-Bt treatment. Moreover, mortality of hosts in the combination of 12% Bt (DP99B) and parasitism was signiWcantly lower than in parasitism alone. The performance of parasitoids is inXuenced by both behavioral and physiological factors. Physiological aVects of transgenic plants on the third trophic level can include both direct toxic eVects of the heterogeneous protein and indirect eVects due to changes in host quality (Schuler et al., 2001). In this study, transgenic cotton had a negative eVect on the development of M. mediator, with the weight of H. armigera larvae fed Bt-diets being signiWcantly less than that of larvae fed the control diet (Fig. 1). Because host size is known to inXuence the development of parasitoids (Harvey et al., 1999; Liu et al., 2004b), we believe the observed eVects were most probably host-quality mediated rather than the direct eVects of the Bt cotton. Chilcutt and Tabashnik (1997) reported that interactions between the parasitoid Cotesia plutellae and Bt depended on host phenotype of Bt-resistance. In susceptible hosts of diamondback moth (Plutella xylestella), the parasitoid did not aVect the performance of Bt, with the pathogen having a signiWcant eVect on the parasitoid. In moderately resistant hosts, the interaction between Bt and the parasitoid had a roughly equal negative impact on each other, with no interaction occurring in highly Bt-resistant hosts. In this study, a highly susceptible host, H. armigera larvae, was used. Indeed, some cases of long-term selection in the laboratory indicate that H. armigera has the potential to develop resistance to Bt transgenic cotton and Bt protein (Liang et al., 2000; Meng et al., 2003; Zhao et al., 1998). Similar studies using Bt-resistant H. armigera larvae in the future will help to understand interactions of Bt cotton and parasitoids. Acknowledgments We are grateful to Yong Zhong and Jie Dong for technical assistance. We also thank Richard Dawson and Hilda Collins for helpful comments on an earlier version of the manuscript. This research was funded by “973” projects (G2000016209) and the State Key Research Programs of Ministry of Science and Technology (2001BA507A-11-02) of PR China. References Akhurst, R.J., James, W., Bird, L.J., Beard, C., 2003. Resistance to the Cry1Ac
-endotoxin of Bacillus thuringiensis in the cotton bollworm, Helicoverpa armigera (Lepidoptera: Noctuidae). J. Econ. Entomol. 96 (4), 1290–1299. Atwood, D.W., Young, S.Y., Kring, T.J., 1997. Development of Cotesia marginiventris (Hymentptera: Braconidae) in tobacco budworm (Lepi-

X. Liu et al. / Biological Control 35 (2005) 134–141 doptera: Nocuidae) larvae treated with Bacillus thuringiensis and thiodicarb. J. Econ. Entomol. 90, 751–756. Atwood, D.W., Young, S.Y., Kring, T.J., 1998. Mortality of tobacco budworm larvae (Lepidoptera: Noctuidae) and emergence of Cotesia marginiventris (Hymenoptera: Braconidae) exposed to Bacillus thuringiensis and thiodicarb alone and in combination. J. Entomol. Sci. 33, 136–141. Baur, M.E., Boethel, D.J., 2003. EVect of Bt-cotton expressing Cry1A(c) on the survival and fecundity of two hymenoptera parasitoids (Braconidae: Encyrtidae) in the laboratory. Biol. Control 26, 325–332. Bell, A.B., Fitches, E.C., Down, R.E., Ford, L., Marris, G.C., Edwards, J.P., Gatehouse, J.A., Gatehouse, A.M.R., 2001. EVect of dietary cowpea trypsin inhibitor (CpTI) on the growth and development of the tomato moth Lacanodia oleracea (Lepidoptera: Noctuidae) and on the success of the gregarious ectoparasitoid Eulophus pennicornis (Hymenoptera: Eulophidae). Pest Manag Sci. 57, 57–65. Blumberg, D., Navon, A., Goldenberg, S.K.S., Ferkovich, S.M., 1997. Interactions among Helicoverpa armigera (Lepidoptera: Noctuidae), its larval endoparasitoid Microplitis croceipes (Hymenoptera: Braconidae), and Bacillus thuringiensis. J. Econ. Entomol. 90 (5), 1181–1186. Chilcutt, C.F., Tabashnik, B.E., 1997. Host-mediated competition between the pathogen Bacillus thuringiensis and the parasitoid Cotesia plutellae of the diamondback moth (Lepidoptera: Plutellidae). Environ. Entomol. 26 (1), 38–45. Cui, J.J., Xia, J.Y., 1999. EVects of transgenic Bt cotton on the population dynamic of natural enemies. Acta Gossypii Sin. 11 (2), 84–91. Deng, S.D., Xu, J., Zhang, Q.W., Zhou, S.W., Xu, G.J., 2003. Resistance of transgenic Bt cotton to the cotton bollworm in cotton Welds in Hubei Province. Acta Entomol. Sin. 46 (5), 584–590. Fan, X.L., Rui, C.H., Xu, C.R., Meng, X.Q., Guo, S.D., Zhao, J.-Z., 2001. Resistance of transgenic cottons expressing Bt and CpTI insecticidal protein genes to Helicoverpa armigera. Acta Entomol. Sin. 44, 582–585. Greenplate, J.T., 1999. QuantiWcation of Bacillus thuringiensis insect control protein Cry1Ac over time in Bollgard cotton fruit and terminals. J. Econ. Entomol. 92 (6), 1377–1383. Guo, S.D., Cui, H., Xia, L., Wu, D., Ni, W., Zhang, Z., Zhang, B., Xu, Y., 1999. Development of bivalent insect–resistant transgenic cotton plants. Sci. Agric. Sin. 32 (4), 1–7. Harvey, J.A., Jervis, M.A., Gols, R., Jiang, N.Q., Vet, L.E.M., 1999. Development of the parasitoid, Cotesia rubecula (Hymenoptera: Braconidae) in Pieris rapae and Pieris brassicae (Lepidoptera: Pieridae): evidence for host regulation. J. Insect Physiol. 45, 173–182. Jackson, R.E., Bradley Jr., J.R, Van Duyn, J.W., Gould, F., 2004. Comparative production of Helicoverpa zea (Lepidoptera: Noctuidae) from transgenic cotton expressing either one or two Bacillus thuringiensis proteins with and without insecticide oversprays. J. Econ. Entomol. 97, 1719–1725. James, C., 2004. International service for the acquisition of agri-biotech application (ISAAA). Brief No. 28, ISAAA, Ithaca, NY. Johnson, M.T., Gould, F., 1992. Interaction of genetically engineered host plant resistance and natural enemies of Heliothis virescens (Lepidoptera: Noctuidae) in tobacco. Environ. Entomol. 21 (3), 586–597. Johnson, M.T., 1997. Interaction of resistant plants and wasp parasitoids of tobacco budworm (Lepidoptera: Noctuidae). Environ. Entomol. 26 (2), 207–214. Liang, G.M., Tan, W.J., Guo, Y.Y., 2000. Study on screening and inheritance mode of resistance to Bt transgenic cotton in H. armigera. Acta Entomol. Sin. 43, 57–62. Liu, X.X., Zhang, Q.W., Cai, Q.N., Li, J.C., Dong, J., 2004a. EVect of Bt protein on development of diVerent strains of the cotton bollworm, Helicoverpa armigera (Hübner) and the parasitoid, Microplitis mediator (Haliday). Acta Entomol. Sin. 47 (4), 461–466. Liu, X.X., Zhang, Q.W., Li, J.C., Xu, J., 2004b. EVects of host size on oviposition and development of the endoparasitoid, Microplitis mediator Haliday. Chin. J. Biol. Control 20 (2), 110–113. Liu, X.X., Zhang, Q.W., Zhao, J.-Z., Cai, Q.N., Xu, H.L., Li, J.C., 2005a. EVects of the Cry1Ac toxin of Bacillus thuringiensis on Microplitis

141

mediator (Haliday) (Hymenoptera: Braconidae), a parasitoid of the cotton bollworm, Helicoverpa armigera (Hübner). Entomol. Exp. Appl. 114, 205–213. Liu, X.X., Sun, C.G., Zhang, Q.W., 2005b. EVects of transgenic Cry1A + CpTI cotton and Cry1Ac toxin on the parasitoid, Campoketis chlorideae (Hymenoptera: Ichneumonidae). Insect Sci. 12, 101–107. Lu, M.G., Rui, C.H., Zhao, J.-Z., Jian, G.L., Fan, X.L., Gao, X.W., 2004. Selection and heritability of resistance to Bacillus thuringiensis subsp kurstaki and transgenic cotton in Helicoverpa armigera (Lepidoptera: Noctuidae). Pest Manag. Sci. 60, 887–893. Meng, F.X., Shen, J.L., Zhu, Z.P., 2003. Temporal-spatial variation in eYcacy of Bt cotton leaves against Helicoverpa armigera (Hübner) and eVect of weather conditions. Acta Entomol. Sin. 46 (3), 299– 304. Pray, C., Huang, J., Qiao, F., 2001. Impact of Bt cotton in China. World Dev. 29, 813–825. Ren, L., Yang, Y.Z., Li, X., Miao, L., Yu, Y.S., Qin, Q.L., 2004. Impact of transgenic Cry1A plus CpTI cotton on Helicoverpa armigera (Lepidoptera: Noctuidae) and its two endoparasitoid wasps Microplitis mediator (Hymenoptera: Braconidae) and Campoletis chlorideae (Hymenoptera: Ichneumonidae). Acta Entomol. Sin. 47 (1), 1–7. Salama, H.S., El-Moursy, A., Zaki, F.N., Aboul-Ela, R., Abdel-Razek, 1991. Parasites and predator of the meal moth Plodia interpunctella Hbn. as aVected by Bacillus thuringiensis Berl. J. Appl. Entomol. 112, 244–253. Schuler, T.H., Potting, R.P.J., Denholm, I., Poppy, G.M., 1999. Parasitoid behaviour and Bt plants. Nature 400, 825–826. Schuler, T.H., Denholm, I., Jouanin, L., Clark, S.J., Clark, A.J., Poppy, G.M., 2001. Population-scale laboratory studies of the eVect of transgenic plants on non-target insects. Mol. Ecol. 10, 1845–1853. Shelton, A.M., Zhao, J.-Z., Roush, R.T., 2002. Economic, ecological, food safety, and social consequences of the deployment of Bt transgenic plants. Annu. Rev. Entomol. 47, 845–881. SPSS 1998. SPSS User’s Guide, (SPSS, Chicago, Illinois). Wan, P., Wu, K., Huang, M., Wu, J., 2004. Seasonal pattern of infestation by pink bollworm Pectinophora gossypiella (Saunders) in Weld plots of Bt transgenic cotton in the Yangtze River valley of China. Crop Prot. 43, 463–467. Wang, D.A., Nan, L.Z., Sun, X., Li, X.Z., 1984. Study on a bionomics of Microplitis spp, larval parasitic wasp of Helicoverpa armigera. Nat. Enemies Insect 6, 211–218. Weseloh, R.M., Andreadis, T.G., 1982. Possible mechanism for synergism between Bacillus thuringiensis and the gypsy moth (Lepidoptera: Lymantriidae) parasitoid, Apanteles melanoscelus (Hymenoptera: Braconidae). Ann. Entomol. Soc. Am. 75, 435–438. Wu, K., Liang, G., Guo, Y., 1997. Phoxim resistance in Helicoverpa armigera (Lepidoptera: Noctuidae) in China. J. Econ. Entomol. 90, 868– 872. Zhang, J.H., Wang, C.Z., Qin, J.D., Guo, S.D., 2004. Feeding behaviour of Helicoverpa armigera larvae on insect-resistance transgenic cotton and non-transgenic cotton. J. Appl. Entomol. 128, 218–225. Zhao, J.-Z., Cao, J., Li, Y.X., Collins, H.L., Roush, R.T., Earle, E.D., Shelton, A.M., 2003. Transgenic plants expressing two Bacillus thuringiensis toxins delay insect resistance evolution. Nat. Biotechnol. 21, 1493– 1497. Zhao, J.-Z., Zhao, K.J., Lu, M.G., Fan, X.L., Guo, S.D., 1998. Interactions between transgenic Bt cotton and Helicoverpa armigera in North China. Sci. Agric. Sin. 31 (5), 1–6. Zhao, J.-Z., Zhao, K.J., Fan, X.L., Lu, M.G., Rui, C.H., Zhang, H.Z., Guo, S.D., 2000. Comparison of insecticidal activity of Bt cotton lines both developed in China and USA against Helicoverpa armigera. Sci. Agric. Sin. 33 (5), 100–102. Zhao, J.-Z., Rui, C.H., 2004. Insect resistance management for transgenic Bt cotton. In: Jia, S.R., Guo, S.D., An, D.C., Xia, G.X. (Eds.), Transgenic Cotton. Science Press, Beijing/New York, pp. 184–204.