INSECTICIDE RESISTANCE AND RESISTANCE MANAGEMENT
Frequency of Alleles Conferring Resistance to the Bt Toxins Cry1Ac and Cry2Ab in Australian Populations of Helicoverpa armigera (Lepidoptera: Noctuidae) R. J. MAHON,1,2 K. M. OLSEN,1 S. DOWNES,3
AND
S. ADDISON4
J. Econ. Entomol. 100(6): 1844Ð1853 (2007)
ABSTRACT Helicoverpa armigera (Hu¨ bner) (Lepidoptera: Noctuidae) is an important lepidopteran pest of cotton (Gossypium spp.) in Australia and the Old World. From 2002, F2 screens were used to examine the frequency of resistance alleles in Australian populations of H. armigera to Bacillus thuringiensis (Bt) Cry1Ac and Cry2Ab, the two insecticidal proteins present in the transgenic cotton Bollgard II. At that time, Ingard (expressing Cry1Ac) cotton had been grown in Australia for seven seasons, and Bollgard II was about to be commercially released. The principal objective of our study was to determine whether sustained exposure caused an elevated frequency of alleles conferring resistance to Cry1Ac in a species with a track record of evolving resistance to conventional insecticides. No major alleles conferring resistance to Cry1Ac were found. The frequency of resistance alleles for Cry1Ac was ⬍0.0003, with a 95% credibility interval between 0 and 0.0009. In contrast, alleles conferring resistance to Cry2Ab were found at a frequency of 0.0033 (0.0017, 0.0055). The Þrst isolation of this allele was found before the widespread deployment of Bollgard II. For both toxins the experiment-wise detection probability was 94.4%. Our results suggest that alleles conferring resistance to Cry1Ac are rare and that a relatively high baseline frequency of alleles conferring resistance to Cry2Ab existed before the introduction of Bt cotton containing this toxin. KEY WORDS Bt resistance, F2 screen, Cry1Ac, allele frequency, Helicoverpa armigera
Concern about the evolution of resistance by insect pest species to Bacillus thuringiensis (Bt) toxins preceded the Þrst deployment of Bt crops (McGaughey and Whalon 1992, Tabashnik 1994, Gould 1998). However, despite insect pests being exposed to Bt crops for up to 10 yr, and the rapidly increasing area planted to these products, to date there is no report of Þeld resistance to Bt crops (Fox 2003, Bates et al. 2005). The lepidopteran Helicoverpa armigera (Hu¨ bner) (Lepidoptera: Noctuidae) is distributed throughout the Old World where it is the major pest on numerous crops, particularly cotton (Gossypium spp.). It poses the principal threat to transgenic cotton in Australia, India, and China (Forrester et al. 1993, Wu and Guo 2005), because of its recidivist nature in evolving resistance to conventional insecticides. The capacity of this species to develop resistance to a Bt toxin (Cry1Ac) is evident from experiments in Australia (Akhurst et al. 2003), India (Kranthi et al. 2000), and China (Fan et al. 2000, Liang et al. 2000) where Þeldcollected populations challenged by selection in the laboratory developed high levels of resistance. 1
CSIRO Entomology, GPO Box 1700 Canberra, ACT 2601 Australia. Corresponding author, e-mail:
[email protected]. CSIRO Entomology, ACRI, Locked Bag 59, Narrabri 2390 Australia. 4 Monsanto Australia Ltd, PO Box 92, Harlaxton, Qld 4350 Australia. 2 3
From 1996 to 2003, the Australian cotton industry deployed Ingard cotton (Bollgard elsewhere) expressing the cry1Ac gene from B. thuringiensis Berliner variety kurstaki. Ingard provided considerable protection for cotton crops against the major insect pests H. armigera and Helicoverpa punctigera (Wallengren). However, the titer of Cry1Ac toxin in the plant decreased in mature cotton partly due to the decline in translation of the cry gene (Olsen et al. 2005). This decline, coupled with the innate tolerance of Helicoverpa species to Bt toxins (Liao et al. 2002), permitted some late season survival of larvae that were susceptible to Bt (Fitt et al. 1998, Fitt 2004). As a consequence, maturing crops often required applications of conventional insecticides to limit larval damage. During the deployment of Ingard in Australia it was considered important to retain susceptibility of H. armigera to Cry1Ac because a second generation of transgenic crops was anticipated that would express this toxin in a pyramid with another Cry toxin. Because the longevity of a two-toxin pyramid depends on the frequency of alleles that confer resistance to either toxin at the time of deployment (Roush 1998), the industry adopted a conservative resistance management plan (RMP) for Ingard to “protect” Cry1Ac that included restricting the area planted to 30% of the total crop.
0022-0493/07/1844Ð1853$04.00/0 䉷 2007 Entomological Society of America
December 2007
MAHON ET AL.: RESISTANCE TO BT TOXINS IN AUSTRALIAN H. armigera
At the end of the 2003Ð2004 cotton growing season, Ingard cotton was withdrawn from the Australian marketplace and replaced by Bollgard II, which expresses both cry1Ac and cry2Ab. Sequence information indicates these genes are distantly related and the toxins they encode do not share a common binding site (Ho¨ fte and Whiteley 1989, English et al. 1994, Liao et al. 2005). Consequently, it is thought unlikely that a single mechanism could confer resistance to both toxins. As a result, the RMP for Bollgard II was relaxed to allow up to 95% of a property to be planted to this crop (Farrell 2006). As in Ingard, the expression of Cry1Ac in Bollgard II decays in older plants, but, although the expression of Cry2Ab also declines over time, levels of Cry2Ab remain consistently high throughout the season (Greenplate et al. 2003). Therefore, Bollgard II provides good control of H. armigera all season. It has been widely adopted in Australia with 71 and 80% of the cotton plantings in 2004 Ð2005 and 2005Ð2006, respectively, being made up of this product. In 1994, the Australian cotton industry established an ongoing monitoring program to detect the early onset of resistance by H. armigera to Bt crops (Downes et al. 2007). From 1994 until 2002, a monitoring method that screens Þeld-collected individuals at a discriminating concentration of MPV containing Cry1Ac (hereafter F0 screens) was used exclusively. However, this approach largely precludes adaptive resistance management if the resistance is recessive, because only large changes in resistance frequencies can be detected. In 2002, an F2 screen (Andow and Alstad 1998) for rare resistance alleles was incorporated into the monitoring program to improve the sensitivity of estimates of resistant allele frequencies and the ability to identify and recover resistance alleles from natural populations. This timing enabled the existing frequency of alleles conferring resistance to Cry1Ac, and the baseline frequency of alleles conferring resistance to Cry2Ab, to be established before the introduction of Bollgard II. Herein, we present data collected from F2 screens on the frequency of alleles conferring resistance to Cry toxins in Australian Þeld populations of H. armigera. The Þrst data were gathered between 2002 and 2004 in the Commonwealth ScientiÞc and Industrial Research Organization Entomology laboratories in Canberra, Australian Capital Territory. Subsequently, the technique was used in the Commonwealth ScientiÞc and Industrial Research Organization Entomology laboratory at the Australian Cotton Research Institute (ACRI) in Narrabri, New South Wales (NSW) and the Monsanto Australia laboratory in Toowoomba, Queensland. Our focus was on detecting alleles conferring resistance to either Cry1Ac or Cry2Ab, the two toxins that are present in Bollgard II. Materials and Methods Standardization of Techniques. The methods and equipment used for setting up and scoring assays, and for rearing insects at contributing laboratories were as uniform as we could achieve. This was ensured by
1845
sourcing materials from the same supplier and by having one person train all personnel involved when implementing a standardized set of techniques. Any differences in the techniques are detailed below, and they are unlikely to alter the ability to detect resistance. Insect Sampling. H. armigera were sampled from all of the major cotton regions in eastern Australia as well as close to the towns of Burdekin in Queensland and Broome in Western Australia (Fig. 1; Table 1). The majority of samples were collected as eggs from non-Bt and Bt-cotton and also from other crops that are hosts to Helicoverpa species. Mostly, we sampled one egg per leaf to reduce the possibility of testing more than one individual from the same female. Usually, collections were made after reports from growers or consultants that eggs were present in their crop, rather than by random sampling throughout a region. Variation in the samples among valleys and throughout the seasons was largely dictated by ßuctuations in population abundance and the availability of dedicated egg collectors. We also sampled a small number of live larvae that were found by consultants or growers on Bollgard II cotton. Rearing Methods. The rearing methods used to maintain H. armigera were the standard procedures described by Teakle and Jensen (1985) with the following exceptions related to the formulation of the diet. In Canberra, soyabean ßour was substituted with chickpea ßour, and at all locations 0.08% propionic acid was substituted for formalin. Rearing trays were covered and heat-sealed by a perforated lid. Moths were provided with a small pot of 4% honey/sugar solution that was fed through a cotton wick and housed in containers that were open at the top and covered with diaper liners or polyester organza cloth secured around their lip. F2 Screening Procedure. F2 tests generate isofemale lines that produce a proportion (1/16) of individuals that are homozygous for haplotypes present in their Þeld-derived parents. The methods for the F2 screens used herein are described by Andow and Alstad (1998). Parental Generation. Field-collected eggs or larvae were reared individually to adults in the laboratory. On maturation, pupae were collected, washed, sexed, and set up in cages that contained pupae of the same sex from the same valley. Emerged adult moths (one male and one female) from the same valley were then placed into containers (350 Ð750 ml) as single pairs. The use of single pairs (hereafter P1 lines) ensured that four haplotypes were tested for the presence of resistant alleles. This technique was preferred over testing Þeld-mated females because multiple mating occurs in this species (Kvedaras et al. 2000), which complicates the interpretation of data. F1 Generation. Containers housing the single pairs were checked at least every 3 d, and liners containing fertile eggs were collected and stored at 6 Ð9⬚C until the female had ceased ovipositing or at least 100 eggs were collected. At this time, the liners were incubated at 25⬚C, which promoted hatching in ⬇3 d. We aimed
1846
JOURNAL OF ECONOMIC ENTOMOLOGY
Vol. 100, no. 6
Fig. 1. Map of Australia depicting the major cotton growing regions, which correspond to those listed in Table 2.
Table 1. The number of isofemale lines progressing through to the F2 screen
Lab
Yr
BM 2002Ð2003 2003Ð2004 2004Ð2005 2005Ð2006 NB 2002Ð2003 2003Ð2004 2004Ð2005 2005Ð2006 TW 2002Ð2003 2003Ð2004 2004Ð2005 2005Ð2006 Total
P1 lines started 322 438 0 0 0 0 391 659 0 442 261 491 3,004
F1 generation
F2 screen
Larvae Moths Cry1Ac Cry2Ab produced produced 49 107
34 87
29 61
28 62
120 242
92 234
91 223
92 223
194 126 158 996
190 124 146 907
190 100 132 826
190 100 132 827
Data are presented separately for each laboratory (BM, Black Mountain; NB, Narrabri; TW, Toowoomba) and year, but they have been combined for all locations and crops (i.e., conventional cotton, Bollgard II cotton, mungbean, sunßowers, chick pea, maize, and pigeon pea).
to rear 135 neonates from each pair individually as isofemale lines. On maturation, pupae were collected, washed, and sexed, and equivalent numbers of males and females were placed in a 5-liter container and allowed to sib-mate in bulk. F2 Generation. Eggs were collected daily and stored at 6 Ð9⬚C. When at least 300 eggs had accumulated over an interval of at least 5 d, they were removed from the liners or cloths by washing in a 0.005% solution of household bleach, Þltered onto a paper disc with a suction funnel, and placed at 25⬚C to hatch. All F2 lines were produced from at least 15 F1 male and 15 F1 female moths. Bt-Susceptible Laboratory Strains. The general laboratory strains used in our assays are susceptible to both Cry1Ac and Cry2Ab toxins. This susceptibility was monitored regularly. In the Canberra and ACRI tests, the susceptible strain was used during every screen to verify that a correctly administered discriminating concentration of toxin-containing material was applied. The susceptible strain used in Canberra (designated GR) has been in culture since the mid-1980s, and it is
December 2007
MAHON ET AL.: RESISTANCE TO BT TOXINS IN AUSTRALIAN H. armigera
derived from material collected from cotton Þelds in the Namoi Valley, northern NSW Australia. On occasion, it has been supplemented with additional collections from the same area. The susceptible strain used in Narrabri (designated ANGR) has been in culture since 1998, and it is derived from GR and a strain designated ANO2 that is made up of material collected from the Namoi Valley. The susceptible strain used in Toowoomba (designated LC) has been in culture since 2002, and it is derived from Þeld collections from throughout the cotton growing areas and has been regularly supplemented with additional collections. F2 Screen Assay. Assays were conducted in 45-well (2.7-cm2) trays that contained ⬇2 ml of rearing diet that was overlaid with an aqueous solution of toxin at the selected discriminating concentration and allowed to air dry. Toxin concentrations were calculated as micrograms per square centimeter of diet surface. After the addition of one neonate larvae per well, trays were heat sealed and maintained at 25⬚C and 45Ð55% RH. We aimed to expose 90 neonate larvae (two 45well trays) to each toxin for each isofemale line. After 7 d, the larvae were scored as being alive (exhibiting normal movement) or dead (dead, moribund, uncoordinated movement), and the growth stage (instar) of all survivors was recorded. Toxins. Cry1Ac. The work performed at all laboratories used Cry1Ac toxin produced by the HD-73 strain of B. thuringiensis variety kurstaki (producing only the Cry1Ac toxin and spores). Mass production via fermentation of HD-73 was performed by Genesearch (Brisbane, Australia) with a resulting spore/ crystal mix. The pellets produced were resuspended and washed three times before use. The extract was used without activating the toxin by trypsin treatment. In all cases, the discriminating concentration for Cry1Ac was 0.25 g/cm2 Cry1Ac delivered in a 50 l/well solution. After 7 d, this concentration killed 95.7 ⫾ 1.8% of GR (n ⫽ 628 larvae in 10 assays conducted over 7 d), and no surviving larvae grew beyond second instar. Cry2Ab. In the 2002Ð2003 season (in Canberra), freshly ground leaves of the experimental transgenic cotton variety Sicala V-2X that expresses the cry2Ab gene was used as a source of Cry2Ab toxin. The cry2Ab gene construct is the property of Monsanto Company (St. Louis, MO). Fresh material from laboratorygrown plants was ground in water until only Þne particles were present, and this slurry was presented to larvae as a surface contaminant. The application rate was 10 mg/cm2. Toxin in the leaf was calibrated using an enzyme-linked immunosorbent assay (ELISA) method on aliquots of leaf material after freeze drying and homogenization. Protein extraction and ELISA methods are detailed in Holt et al. (2002) except that Cry2Ab antibodies and controls were substituted for their Cry1Ac equivalents. The aliquot of 10 mg/cm2 was estimated by ELISA to contain 2 g/cm2 Cry2Ab toxin. This screening concentration was selected on the basis of the mortality it induced and also the suppression of growth among the small proportion of
1847
larvae that survived. After 7 d, the concentration of toxin caused 96 ⫾ 1.1% (n ⫽ 744 in 15 assays conducted over 7 d) of GR neonate larvae to die or, if they did survive, they did not grow beyond Þrst instar. No larvae reached third instar. From 2004 to the present, dried and ground corn leaf material was used as a source of Cry2Ab toxin at all laboratories. This corn powder was provided by Monsanto (United States) as a lyophilized Zea mays L. leaf powder containing the transgenically expressed B. thuringiensis crystal protein Cry2Ab2, at a concentration of 6 mg/g powder. The powder, used at 1 g/cm2, was slightly more efÞcacious against GR than the cotton leaf material, because after 7 days 99.6 ⫾ 0.4% of larvae tested either died or did not grow beyond Þrst instar and mortality was 96 ⫾ 1.1% (n ⫽ 286 larvae over six assays conducted over 7 d). No larvae reached third instar. Isolating Resistant Alleles. Surplus F2 eggs or larvae were retained pending the result of the assays. These were discarded if both the Cry1Ac and Cry2Ab screens did not show signiÞcant levels of survival. However, if at least 6% of larvae survived the discriminating concentration for either toxin as at least second instars, or one or more larvae reached the third instar, the surplus material was reared to adults along with all survivors of the F2 screen. The progeny of the larvae that survived the discriminating concentration of toxin were mated among themselves or crossed to the untreated cohort. To conÞrm that the line carried an allele conferring resistance to the toxin, the F3 generation was retested at the same concentration set for the F2 screen. In all cases, retested lines again showed evidence of resistance and they were retained. At this stage, the F2 isofemale line was deemed to have scored positive for an allele conferring resistance to the toxin. Statistical Analysis. There are two main statistical calculations necessary when doing an F2 screen: the expected frequency of resistance alleles in the wild population (E[pR]) with credibility intervals and the probability of missing a resistance allele that was present in nature (PNo). Here, we are more interested in the actual allele frequency of our sampled population than in the average allele frequency of a large number of similar populations. Thus, we use Bayesian inference (Brunk 1975) to estimate the expected allelic frequency and the 95% credibility intervals. Andow and Alstad (1998) provide a thorough justiÞcation for using this statistical approach in relation to F2 screens. It is possible to erroneously conclude that an isofemale line carries only susceptible alleles when it actually had resistance alleles. It can occur, for example, when a line starts with a resistance allele, but the allele is lost or reduced to a low frequency in the F1 or F2 generation and becomes difÞcult to detect. We followed the methods developed by Andow and Alstad (1998) and Stodola and Andow (2004) for calculating PNo. This probability is explicit for recessive alleles and depends on the number of F1 males and F1 females that contribute to the F2 generation, the number of F2
1848
JOURNAL OF ECONOMIC ENTOMOLOGY
offspring screened per F1 female, and the nonscreen mortality of F2 larvae. Our counts of F1 parents only include those that contributed to the F2 screen. In 2003Ð2004, 2004 Ð2005, and 2005Ð2006 for Toowoomba and Narrabri, nonscreen mortality of larvae was measured for each isofemale line. We reared a random selection of 45 F2 neonates on artiÞcial diet in the same room as their siblings that were undergoing the F2 screens. After 7 d, we counted the dead individuals plus those that did not reach at least third instar, and we divided this value by the total number of individuals tested. In 2002Ð2003 and 2003Ð2004 for Canberra and 2004 Ð2005 for Narrabri, we used the average value calculated across all isofemale lines for the four cases above as an estimate of nonscreen mortality (mean ⫾ SD, 0.30 ⫾ 0.20). At least 90% of our screens against Cry1Ac and Cry2Ab contained between 80 and 90 offspring (range, 39 Ð 115; mean, 86.3). For each isofemale line the actual numbers of offspring tested were included in our calculation of the number of F2 offspring screened per F1 female. We used goodness-of-Þt statistics (G tests; Sokal and Rohlf 1969) to compare the frequencies of survivors within isofemale lines that scored positive for resistance relative to the frequencies of survivors that would be expected under a hypothesis that the resistance was recessive. If resistance is completely recessive, we would expect 6.25% (1/16) of larvae to develop to at least the third instar at the time of screening our assays. Results Progression through F2 Experiment. From the 2002Ð2003 season until the most recent season (2005Ð 2006), 3,304 alleles (826 isofemale lines) were screened against Cry1Ac and 3,308 alleles (827 isofemale lines) were screened against Cry2Ab by using the F2 test (Table 1). Table 1 lists a detailed breakdown of the number of P1 lines that were started, and the number of lines that passed through a deÞned stage, in the contributing laboratories over the 3-yr study. Of the 3,004 lines started, only 27% were successfully screened. Screening success varied among laboratories from ⬇13% for the initial work carried out at Black Mountain to 30 Ð35% for Narrabri and Toowoomba, but it was fairly consistent in the same laboratory among years (⬍10% variation in all cases). In all laboratories, most unsuccessful lines were lost during the P1 generation (Table 1) through a failure to mate, mating that resulted in nonviable eggs, or mating that resulted in few viable eggs. The majority (91%) of F1 larvae successfully emerged as F1 moths and met our criteria of at least 15 males and 15 females contributing to the F2 generation screen. Frequency of Alleles Conferring Resistance. The combined data from all laboratories, years, and collecting sites show that none of the 826 isofemale lines examined for resistance to Cry1Ac scored positive (Table 2). Based on these data, the estimated R frequency for alleles conferring resistance to Cry1Ac in
Vol. 100, no. 6
Australia is ⬍0.0003 with a 95% credibility interval (CI) between 0 and 0.0009. In contrast, 10 of the 827 isofemale lines examined for resistance to Cry2Ab scored positive (Table 2). Based on these data, the estimated R frequency for alleles conferring resistance to Cry2Ab in Australia is 0.0033 with a 95% CI between 0.0017 and 0.0055. When data were pooled across years, the frequency of detecting isofemale lines that scored positive for Cry2Ab resistance did not differ signiÞcantly among laboratories (Fisher exact test 2 ⫽ 4.1, df ⫽ 2, P ⫽ 0.13; Tables 1 and 2). Similarly, when data were pooled across laboratories, the frequency of detecting isofemale lines that scored positive for Cry2Ab resistance did not differ signiÞcantly among years (Fisher exact test 2 ⫽ 5.6, df ⫽ 3, P ⫽ 0.13; Tables 1 and 2). It is informative that the 10 isolations of Cry2Ab resistance alleles were from the four sampling locations with the greatest screening effort, namely, the Namoi (3/300 ⫽ 0.010), Darling Downs (3/167 ⫽ 0.018), Macquarie (2/82 ⫽ 0.024), and Gwydir (2/ 73 ⫽ 0.027) valleys (Table 2). Probability of False Negatives. There was ⬍1% probability of having missed a resistance allele in 775 of the 826 lines screened for Cry1Ac and 827 lines screened for Cry2Ab. The experiment-wise detection probability was 94.4%. Survival in Isofemale Lines That Carried Resistance Alleles. In the 10 lines that were scored positive for the presence of a Cry2Ab resistance allele, ⬎6.25% of larvae developed to at least the third instar at the time of screening our assays (Table 3). In Þve of these lines, the difference between the observed frequencies and expected frequencies is statistically signiÞcant at the P ⬍ 0.05 level (Table 3). A replicated goodness-of-Þt test indicates that the survival frequencies among the isofemale lines scoring positive for Cry2Ab resistance is heterogeneous (2 ⫽ 38.47, df ⫽ 9, P ⬍ 0.001), and overall, the Þt to a 1:15 ratio is poor (2 ⫽ 109.36, df ⫽ 1, P ⬍ 0.001). Discussion The tests used are likely to detect nontrivial forms of resistance to Cry1Ac or Cry2Ab that are determined by a single locus. The F2 screen is not infallible, and some forms of resistance may not be detected, particularly if they are due to more than one locus. However, such forms of resistance develop infrequently in the Þeld (Roush and McKenzie 1987, McKenzie et al. 1992), but they remain possible outcomes of selection (Groeters and Tabashnik 2000). In this study, the probability of obtaining false negatives (estimated using PNo) by using the aforementioned testing regime is very low; therefore, we consider that our estimates of the frequency of resistance in H. armigera to both toxins are robust. Of the 3,304 single pairs of moths set up in this study, only 826 (27%) successfully progressed to the stage where F2 neonate larvae were exposed to toxin. It is appropriate to consider whether the successful single pairs represent a biased sample that could inßuence
December 2007 Table 2.
MAHON ET AL.: RESISTANCE TO BT TOXINS IN AUSTRALIAN H. armigera
1849
The number of lines in the F2 screen that scored positive for carrying a resistance allele to Cry1Ac or Cry2Ab Cry2Ab F2 screen
Cry1Ac F2 screen Yr
Locality
2002Ð2003
Macquarie Namoi Season total Macquarie Namoi Gwydir Emerald MacIntyre Darling Downs St George Season total Macquarie Namoi Gwydir MacIntyre St George Darling Downs Emerald Burdekin Broome Season total Macquarie St George MacIntyre Gwydir Emerald Dirrinbandi Darling Downs Bourke Namoi Burdekin Season total
2003Ð2004
2004Ð2005
2005Ð2006
Tested
Scored positive
Tested
Scored positive
19 10 29 63 49 14 19 36 47 23 251 5 78 7 6 1 52 11 4 27 191 2 12 16 52 33 6 68 1 163 2 355
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
18 10 28 64 49 14 19 36 47 23 252 5 78 7 6 1 53 11 4 27 192 2 12 16 52 33 6 68 1 163 2 355
SP15 0 1 SP202 0 SP566 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 NB271 0 0 NB550, TW661, TW876 0 NB405, NB738, TW1068 0 7
Data are presented separately for each year and locality (as indicated on Fig. 1), but they have been combined for all laboratories and crops (i.e., conventional cotton, Bollgard II cotton, mungbean, sunßowers, chick pea, maize, and pigeon pea). Each isofemale line that scored positive has been designated a code that corresponds to the laboratory in which the screen was conducted (SP, Black Mountain; NB, Narrabri; TW, Toowoomba).
the frequency of detected resistance alleles. Most single pairs that failed to progress to the F2 stage did so because the Þeld-derived moths did not produce fertile F1 offspring. We would only expect a bias in the potential to detect resistance alleles if the tendency to Table 3. Actual and expected numbers of resistant individuals in the initial F2 screens of isofemale lines that scored positive for an allele that conferred resistance to Cry2Ab Larvae ⱖ third instar at day 7 Probability of Isofemale No. Observed Þt to line tested Observed Expected proportion expected no. no. (%) SP15 SP566 SP202 NB271 NB405 NB550 NB738 TW661 TW876 TW1068
115 90 90 87 90 88 87 90 90 90
17 32 14 13 10 17 9 21 6 9
7.19 5.62 5.62 5.44 5.62 5.50 5.43 5.62 5.62 5.62
14.8 35.6 15.6 14.9 11.1 19.3 10.3 23.3 6.7 10.0
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.056 ⬍0.001 0.115 ⬍0.001 0.870 0.142
The expected number assumes that resistance is recessive and thus one in every 16 (6.25%) larvae would be homozygous resistant and thus likely to be at least a third instar at the time of scoring.
mate under laboratory conditions was a pleiotropic effect of the resistance. We are aware of only one study that has addressed this issue for Lepidoptera. In pink bollworm, a maleÕs resistance status affects the chances of obtaining paternity in double-mated females, but it does not affect propensity to mate (Higginson et al. 2005). From 2002 until 2006, we scored 3,304 alleles from Þeld populations of H. armigera for resistance to Cry1Ac and none were positive. From these data, we conclude that Cry1Ac resistance is rare in Þeld populations of H. armigera from Australia, and that it occurs at a frequency of ⬍0.0003. However, laboratory selection yielded a colony (designated BX) that is ⬎400 times resistant to Cry1Ac than a Btsusceptible laboratory strain (Akhurst et al. 2003). Thus, unless a mutation occurred within the colony, the variability within the founding colony was sufÞcient to permit the evolution of resistance. We are conÞdent that if the BX phenotype had been encountered in our F2s as a monogenetic form of resistance, it would have been readily identiÞed because ⬎95% of such individuals would be expected to survive to third instar in our Cry1Ac assays.
1850
JOURNAL OF ECONOMIC ENTOMOLOGY
Dang and Gunning (2002) described a colony of H. armigera (designated ÔSilver strainÕ) established from survivors of a discriminating concentration of the Cry1Ac-containing commercial product MVP during 2000 Ð2001 in Australia, which showed a marked level of resistance to Cry1Ac. However, in 2001 and 2002, we tested individuals of this strain, but we did not Þnd evidence of resistance (R.J.M. and K.O., unpublished data). Gunning et al. (2005) further evaluated the Silver strain and suggested that the resistance is due to a semidominant trait linked to the overexpression of esterase enzymes. The F2 screen used herein is capable of detecting single-gene forms of resistance that are recessive (Andow and Alstad 1998); therefore, forms that exhibit any dominance would be prominent and detected. Assays performed on our own laboratory strains that exhibit a 20 Ð30-fold resistance ratio for Cry1Ac as heterozygotes, demonstrated that heterozygotes that possess low levels of resistance are conspicuous in the F2 screen (R.J.M. and K.O., unpublished data). Thus, we conclude that the form of resistance reported by Gunning et al. (2005) has not been encountered in our sampling. Perhaps the conservative RMP imposed for Ingard was effective in minimizing any increase in frequency of genes conferring resistance to Cry1Ac during the period that it was grown, i.e., 1996 Ð2004. Alternatively, characteristics of Cry1Ac resistance present in the population may not promote its evolution. In the BX colony, and a derived colony (designated IS), resistance is partially dominant (Daly and Olsen 2000), although it is only functionally dominant at certain times on maturing transgenic cotton (Bird and Akhurst 2004, 2005). In addition, the BX colony develops more slowly than a Bt-susceptible strain when it is reared on non-Bt diet in the laboratory (Akhurst et al. 2003), and it suffers other Þtness costs relating to reproduction (Bird and Akhurst 2005). However, these Þtness costs are recessive and are relatively mild compared with the substantial costs incurred by Cry1Ac-resistant pink bollworm Pectinophora gossypiella (Saunders). In this species, resistant genotypes demonstrate reduced larval survival on non-Bt cotton (Carrie` re et al. 2001a), and poor overwintering success (Carrie` re et al. 2001b), which is thought to have played an important role in delaying Þeld resistance in this pest (Tabashnik et al. 2005). To date, no results from F2 screens on H. armigera have been reported from populations beyond Australia. Thus the frequencies of rare resistance alleles in Þeld populations are largely unknown elsewhere even though cry1Ac-containing cotton varieties are grown widely in India and China. F0 screens performed on populations in China suggest that Cry1Ac resistance is rare (Li et al. 2004, Wu et al. 2006), although this method would not detect recessive forms of resistance unless they became so frequent that homozygous individuals became common. The initial frequencies of alleles that confer resistance to new insecticides or Bt toxins have not been well deÞned. In the absence of prior selection usually these frequencies are assumed in the order of 10⫺6
Vol. 100, no. 6
(Gould 1998), although models of the evolution of resistance generally use a “conservative” maximum frequency of 10⫺3 (e.g., Roush 1998). Our screens demonstrate that 10 of 3,308 alleles tested from H. armigera conferred resistance to Cry2Ab. Resistance was isolated from the four most intensively sampled cotton-growing areas in Australia, rather than a particular region. Small sample sizes, and the rarity of isolations of resistance, prevent meaningful statistical comparisons among valleys. However, the widespread origin of the resistance, and its detection before the extensive deployment of Bollgard II, suggests that resistance existed at a relatively high frequency before the introduction of Bollgard II. Two important observations are suggested by the proportion of survivors among the F2 larvae in those isofemale lines that proved to include a resistance allele. First, a larger proportion of larvae survived our screens than expected (1/16) for a recessive form of resistance. This result suggests that some degree of dominance may be present. However, this conclusion is in conßict with research performed with the Þrst isolated Cry2Ab resistant line (designated SP15), which was characterized as recessive (Mahon et al. 2007), suggesting that other explanations also may be possible. Second, the frequency of survival in these isofemale lines is heterogeneous that implies that the resistance mechanisms or their pleiotropic effects may vary among the 10 isolations. Why are alleles conferring resistance to Cry2Ab common? Foliar sprays of DiPel (expressing Cry1Ac, Cry1Aa, Cry1Ab, and Cry2Aa) and MVP (expressing Cry1Ac) have been applied for many years against H. armigera, especially in cotton, but they do not contain the Cry2Ab toxin as a major component. Furthermore, Cry1A toxins present in Dipel are far more potent for Lepidoptera than the minor toxin Cry2Aa. It is possible that soil-borne Bt bacteria that express Cry2Ab selected for resistance. Dust or wet soil particles deposited after rain that contain Bt and/or toxins could lodge on host leaves and be consumed by larvae, or mature larvae entering the soil to pupate could contact the toxin. Examination of Bt collected from various parts of the world (e.g., Mexico, Korea, China, Spain, Costa Rica, and Columbia) shows markedly different relative frequencies of Cry2 and Cry1 toxins (Bravo et al. 1998, Kim 2000, Wang et al. 2003, Martinez et al. 2005, Arrieta and Espinoza 2006, Armengol et al. 2006), and on occasion, Cry2 toxins are more prevalent than Cry1 toxins. However, the relative potency of Cry1 to H. armigera would suggest that if soil-borne Bt is responsible for selection, as long as Þtness costs are recessive, the background frequency of alleles that confer resistance to Cry1Ac would be expected to be elevated relative to that for Cry2Ab resistance. Although genes that confer resistance to Cry1Ac are known to exist in populations of H. armigera, here we show that they are rare. The frequency of mutations in a population is largely determined by the mutation rate, selection coefÞcient, and Þtness costs associated with a mutation (Falconer 1964). Until the mutation becomes
December 2007
MAHON ET AL.: RESISTANCE TO BT TOXINS IN AUSTRALIAN H. armigera
common, only dominant Þtness costs play a signiÞcant role. It is possible that before the introduction of Bollgard II an agent in the environment other than Bt favored selection for alleles conferring resistance to Cry2Ab. If so, it would be difÞcult to isolate this agent, especially now that Bollgard II is widespread. Likewise, there is no obvious means to assess the mutation rate to the resistant state. Thus, the only factor amenable to analysis is Þtness costs associated with the resistance. Such costs are presently being studied (R.J.M. and K.O.). This information is also critical for determining the future rate of evolution of resistance to Cry2Ab by H. armigera. The estimated frequencies of Cry2Ab resistance for each season provide no compelling evidence that Cry2Ab resistance is becoming more common despite the seven isolations made during the most recent (2005Ð2006) season. Practical limitations on the number of alleles that can be screened, combined with the infrequent isolation of resistant alleles, means that the statistical power to detect small changes is poor. However, if a marked change occurs, or a trend develops, the F2 monitoring system is a sensitive means to detect it. In contrast, given the recessive nature of the alleles so far characterized, the more traditional F0 tests used in Australia would not detect an increase in the frequency of rare resistance alleles until they become sufÞciently common to make it futile to use remedial actions (Andow and Ives 2002). Bt resistance in the beetle Chrysomela tremulae F. presents an analogous situation to our Þnding of an unexpectedly high background frequency of resistant alleles in H. armigera. Ge´ nissel et al. (2003) and Wenes et al. (2006) by using F2 screens found that alleles conferring resistance to Cry3Aa in C. tremulae occurred at a frequency of 0.0049. C. tremuale feed only on poplars, are not exposed to Bt foliar sprays, and the work was conducted before the release of transgenic plants. The resistance is recessive and incurs a Þtness cost that probably is not dominant (Augustin et al. 2004). The unexpectedly high frequency of resistance in C. tremulae and H. armigera is in marked contrast to the situation in the European corn borer, Ostrinia nubilalis (Hu¨ bner), in which, despite extensive surveys, no resistance has been identiÞed (Andow et al. 1998, Stodola et al. 2006). If the resistance detected herein in H. armigera proves to be recessive, as shown for SP15 in the laboratory (Mahon et al. 2007), only homozygous resistant insects will be able to tolerate Cry2Ab toxin. With a resistance frequency of 0.0033, homozygote individuals that are capable of surviving on cotton expressing Cry2Ab toxin will be extremely rare at (0.0033)2 or 10 in a million. Bollgard II is grown on ⬎80% of the area devoted to cotton in Australia, and it expresses both cry2Ab and cry1Ac genes. To overcome this technology, any Cry2Ab-resistant homozygote insect also would have to survive the Cry1Ac present in this variety. The frequency of resistance to Cry1Ac is extremely rare (based on the data presented in this study) and Cry2Ab resistant insects are susceptible to Cry1Ac (Mahon et al. 2007). Therefore, the frequency
1851
of insects that are resistant to both toxins would be exceedingly small. However, the titer of both toxins vary markedly in Bollgard II (Cry1Ac range, 0.39 Ð 4.19; Cry2Ab, 4.55Ð33.3 mg/kg fresh weight of leaf material; APVMA 2003), thus potentially providing opportunities for insects resistant to only one toxin. Such opportunities are likely to be more common late in the season for insects that are resistant to Cry2Ab but susceptible to Cry1Ac due to a decline in the Cry1Ac titer (Greenplate et al. 2003). A similar decline in titer enabled some fully susceptible H. armigera to survive on Cry1Ac-expressing Ingard varieties late in the growing season in Australia (Fitt et al. 1998). Thus, it is possible that homozygous Cry2Ab resistant insects will be favored late in the season on mature Bollgard II. Without knowledge of potential Þtness costs associated with Cry2Ab resistance, it is difÞcult to assess the threat that it poses to the longterm efÞcacy of Bollgard II. Bollgard II is presently only grown in Australia, the United States, and India. It is of interest to note that the Old World H. armigera and the New World Helicoverpa zea (Boddie) are closely related and hybrids between the two “species” are possible (Kerkpatrick 1962, Laster and Hardee 1995, Laster and Sheng 1995). Furthermore, sequence studies of the piggyBac elements shared by both H. armigera and H. zea (Zimowska and Handler 2006) support the proposal that these “species” are conspeciÞc (Progue 2004). Because we think the SP15-like polymorphism preexisted selection through the introduction of transgenic cotton, we anticipate that other populations of H. armigera, and perhaps populations of H. zea, may prove to be similarly polymorphic and thereby play a role in the evolution of resistance to Bollgard II in the New World. Acknowledgments K. Stanford, L. Scott, K. Stewart, N. Winters, E. Cuell, R. Isen, C. Mares, B. James, and K. Deaves provided valuable technical assistance. In particular, K. Garsia, T. Parker, and C. Johnston performed the bulk of the technical work in Canberra, Narrabri, and Toowoomba, respectively. We thank L. Rossiter and team (NSW DPI) for coordinating the collection and delivery of eggs while the work was performed at Canberra, and the numerous industry participants and Monsanto Þeld staff that provided alerts and collections of eggs for the screens. Cheryl Mares kindly provided the Sicala V-2X seeds used in the experiment that were produced by Greg Constable of Commonwealth ScientiÞc and Industrial Research Organization Plant Industry and measured the Cry2Ab levels in the leaf material. D. Andow and T. Stodola provided the program used to calculate PNo, calculations of resistance frequencies, and advice on F2 screen statistics. D. Andow provided useful comments on an earlier draft of this manuscript. The Monsanto Company provided the sources of Cry2Ab and R. Akhurst and B. James (Commonwealth ScientiÞc and Industrial Research Organization Entomology) provided the source of Cry1Ac. The work performed in Canberra and Narrabri was funded by the Cotton Research and Development Corporation in Australia, and the Monsanto Company supported the work performed in Toowoomba.
1852
JOURNAL OF ECONOMIC ENTOMOLOGY References Cited
Akhurst, R. J., W. James, L. J. Bird, and C. Beard. 2003. Resistance to the Cry1Ac delta-endotoxin of Bacillus thuringiensis in the cotton bollworm, Helicoverpa armigera (Lepidoptera: Noctuidae). J. Econ. Entomol. 96: 1290Ð 1299. Andow, D. A., and D. N. Alstad. 1998. F2 screen for rare resistance alleles. J. Econ. Entomol. 91: 572Ð578. Andow, D. A., and A. R. Ives. 2002. Monitoring and adaptive resistance management. Ecol. Appl. 12: 1378 Ð1390. Andow, D. A., D. N. Alstad, Y.-H. Pang, P. C. Bolin, and W. D. Hutchison. 1998. Using an F2 screen to search for resistance alleles to Bacillus thuringiensis toxin in European corn borer (Lepidoptera: Crambidae). J. Econ. Entomol. 91: 579 Ð584. [APVMA] Australian Pesticides and Veterinary Medicines Authority. 2003. Evaluation of the new active Bacillus thuringiensis var. kerstaki delta-endotoxins as produced by the Cry1Ac and Cry2Ab genes and their controlling sequences in the new product Bollgard II cotton event 15985. Australian Pesticides and Veterinary Medicines Authority, Canberra, Australia. (http://www.apvma.gov. au/publications/prsbollgard2.pdf). Arrieta, G., and A. M. Espinoza. 2006. Characterization of a Bacillus thuringiensis strain collection from diverse Costa Rican natural ecosystems. Rev. Biol. Trop. 54: 13Ð27. Armengol, G., M. C. Escobar, M. E. Maldonado, and S. Orduz. 2006. Diversity of Columbian strains of Bacillus thuringiensis with insecticidal activity against dipteran and lepidopteran insects. J. Appl. Microbiol. 102: 77Ð 88. Augustin, S., C. Courtin, A. Rejasse, P. Lorme, A. Genissel, and D. Bourguet. 2004. Genetics of resistance to transgenic Bacillus thuringiensis poplars in Chrysomela tremulae (Coleoptera: Chrysomelidae). J. Econ. Entomol. 97: 1058 Ð1064. Bates, S. L., J.-Z. Zhao, R. T. Roush, and A. M. Shelton. 2005. Insect resistance management in GM crops: past, present and future. Nature (Lond.) 23: 57Ð 62. Bird, L. L., and R. J. Akhurst. 2004. The relative Þtness of Þtness of Cry1A-resistant and -susceptible Helicoverpa armigera (Lepidoptera: Noctuidae) on conventional and transgenic cotton. J. Econ. Entomol. 97: 1699 Ð1709. Bird, L. L., and R. J. Akhurst. 2005. Fitness of Cry1A-resistant and -susceptible Helicoverpa armigera (Lepidoptera: Noctuidae) on transgenic cotton with reduced levels of Cry1Ac. J. Econ. Entomol. 98: 1311Ð1319. Bravo, A., S. Sarabia, L. Lopez, H. Ontiveros, C. Abarca, A. Ortiz, M. Ortiz, L. Lina, F. J. Villalobos, G. Pena, et al. 1998. Characterization of cry genes in a Mexican Bacillus thuringiensis strain collection. Appl. Environ. Microbiol. 64: 4965Ð 4972. Brunk, H. D. 1975. An introduction to mathematical statistics. Xerox College Publishing, Lexington, KY. Carrie`re, Y., C. Ellers-Kirk, Y.-B. Liu, M. A. Sims, A. L. Patin, T. J. Dennehy, and B. E. Tabashnik. 2001a. Fitness costs and maternal effects associated with resistance to transgenic cotton in the pink bollworm (Lepidoptera: Gelechiidae). J. Econ. Entomol. 94: 1571Ð1576. Carrie` re, Y., C. Ellers-Kirk, A. L. Patin, M. A. Sims, S. Meyer, Y.-B. Liu, T. J. Dennehy, and B. E. Tabashnik. 2001b. Overwintering cost associated with resistance to transgenic cotton in the pink bollworm (Lepidoptera: Gelechiidae). J. Econ. Entomol. 94: 935Ð941. Daly, J., and K. Olsen. 2000. Genetics of Bt resistance pp185Ð188. In Proceedings of the 10th Australian Cotton Conference, 16 Ð18 August 2000, Brisbane, Australia. Australian Cotton Growers Research Association, Narrabri, Australia.
Vol. 100, no. 6
Dang, H. T., and R. Gunning. 2002. Evidence of the shift in susceptibility to Bacillus thuringiensis delta-endotoxin Cry1Ac in Australian Helicoverpa armigera (Lepidoptera: Noctuidae). Resistant Pest Manag. Newsl. 11: 44 Ð 48. Downes, S. J., R. Mahon, and K. Olsen. 2007. Adaptive resistance management in Australia for Bt-cotton: current status and future challenges. J. Invertebr. Pathol. 95: 208 Ð 213. English, L., H. Loidl Robbins, M. A. von Tersch, C. A. Kulesza, D. Ave, D. Coyle, C. S. Jany, and S. L. Slatin. 1994. Mode of action of CryIIA a Bacillus thuringiensis delta-endotoxin. Insect Biochem. Mol. Biol. 24: 1025Ð 1035. Falconer. D. S. 1964. Introduction to quantitative genetics. Oliver and Boyd, Glasgow, Scotland. Fan, X., J.-Z. Zhao, Y. Fan, and X. Shi. 2000. Inhibition of transgenic Bt plants to the growth of cotton bollworm. Plant Protect. 26: 3Ð5. Farrell, T. 2006. Cotton pest management guide 2006/07. NSW Department of Primary Industries, Orange, NSW, Australia. Fitt, G. P. 2004. Implementation and impact of transgenic Bt cottons in Australia. In Proceedings of the World Cotton Research Conference-3, Cape Town, 9 Ð13 March 2003. Agricultural Research Council, Institute for Industrial Crops, Pretoria, Republic of South Africa. Fitt, G. P., J. C. Daly, C. L. Mares, and K. Olsen. 1998. Changing efÞcacy of transgenic Bt Cotton Ð patterns and consequences, pp. 189 Ð196. In M. P. Zalucki, R.A.I. Drew, and G. G. White [eds.], Pest managementÐfuture consequences, vol. 1. University of Queensland Printery, Brisbane, Australia. Forrester, N. W., M. Cahill, L. Bird, and J. K. Layland. 1993. Management of pyrethroid and endosulfan resistance in Helicoverpa armigera (Lepidoptera: Noctuidae) in Australia. Bull. Entomol. Suppl. Ser. No. 1. 1Ð132. Fox, J. L. 2003. Resistance to Bt toxin surprisingly absent from pests. Nat. Biotechnol. 21: 958 Ð959. Ge´nissel, A., S. Augustin, C. Courtin, G. Pilate, P. Lorme, and D. Bourguet. 2003. Initial frequency of alleles conferring resistance to Bacillus thuringiensis poplar in a Þeld population of Chrysomela tremulae. Proc. R. Soc. Lond. B 270: 791Ð797. Gould, F. 1998. Sustainability of transgenic insecticidal cultivars: integrating pest genetics and ecology. Annu. Rev. Entomol. 43: 701Ð726. Greenplate, J. T., J. W. Mullins, S. R. Penn, A. Dahm, B. J. Reich, J. A. Osborn, P. R. Rahn, L. Ruschke, and Z. W. Shappley. 2003. Partial characterization of cotton plants expressing two toxin proteins from Bacillus thuringiensis: relative toxin contribution, toxin interaction, and resistance management. J. Appl. Entomol. 127: 340 Ð347. Groeters, F. R., and B. E. Tabashnik. 2000. Roles of selection intensity, major genes, and minor genes in evolution of insecticidal resistance. J. Econ. Entomol. 93: 1580 Ð1587. Gunning, R. V., H. T. Dang, F. C. Kemp, I. C. Nicholson, and G. D. Moores. 2005. New resistance mechanism in Helicoverpa armigera threatens transgenic crops expressing Bacillus thuringiensis Cry1Ac toxin. Appl. Environ. Microbiol. 71: 2558 Ð2563. Higginson, D. M., S. Morin, M. E. Nyboer, R. W. Biggs, B. E. Tabashnik, and Y. Carriere. 2005. Evolutionary tradeoffs in insect resistance to Bacillus thuringiensis crops: Þtness cost affecting paternity. Evolution 59: 915Ð920. Holt, H. E., C. Mares, and R. Akhurst. 2002. Determination of the Cry protein content of Bt transgenic cotton. CSIRO Technical Report No. 92. Commonwealth ScientiÞc and Industrial Research Organisation, Canberra, Australia.
December 2007
MAHON ET AL.: RESISTANCE TO BT TOXINS IN AUSTRALIAN H. armigera
Ho¨ fte, H., and H. R. Whiteley. 1989. Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53: 242Ð 255. Kerkpatrick, T. H. 1962. Methods of rearing Heliothis species and attempted crossbreeding. Qld. J. Agric. Sci. 19: 565Ð566. Kim, H.-S. 2000. Comparative study of the frequency, ßagellar serotype, crystal shape, toxicity and cry gene contents of Bacillus thuringiensis from three environments. Curr. Microbiol. 41: 250 Ð256. Kranthi, K. R., S. Kranthi, S. Ali, and S. K. Banerjee. 2000. Resistance to Cry1Ac ␦-endotoxin of Bacillus thuringiensis in a laboratory selected strain of Helicoverpa armigera (Hubner). Curr. Sci. 78: 1001Ð1004. Kvedaras, O. L., P. C. Gregg, and A.P.D. Socorro. 2000. Techniques used to determine the mating behaviour of Helicoverpa armigera (Hubner) (Lepidoptera: Noctuidae) in relation to host plants. Aust. J. Entomol. 39: 188 Ð194. Laster, M. L., and D. D. Hardee. 1995. Intermating compatibility between North American Helicoverpa zea and Heliothis armigera (Lepidoptera, Noctuidae) from Russia. J. Econ. Entomol. 88: 77Ð 80. Laster, M. L., and C. F. Sheng. 1995. Search for hybrid sterility for Helicoverpa zea in crosses between the NorthAmerican Heliothis zea and Helicoverpa armigera (Lepidoptera, Noctuidae) from China. J. Econ. Entomol. 88: 1288 Ð1291. Li, G., K. Wu, F. Gould, H. Feng, Y. He, and Y. Guo. 2004. Frequency of Bt resistance genes in Helicoverpa armigera populations from the Yellow River cotton-farming region of China. Entomol. Exp. Appl. 112: 135Ð143. Liao, C., L. Brooks, S. C. Trowell, and R. J. Akhurst. 2005. Binding of Cry ␦-endotoxins to brush border membrane vesicles of Helicoverpa armigera and Helicoverpa punctigera (Lepidoptera: Noctuidae). Insect Sci. 12: 231Ð240. Liao, C., D. G. Heckel, and R. Akhurst. 2002. Toxicity of Bacillus thuringiensis insecticidal proteins for Helicoverpa armigera and Helicoverpa punctigera (Lepidoptera: Noctuidae), major pests of cotton. J. Invertebr. Pathol. 80: 55Ð 63. Liang, G., W. Tan, and Y. Guo. 2000. Study on screening and inheritance mode of resistance to Bt transgenic cotton in H. armigera. Acta Entomol. Sin. 43 (Suppl.): 57Ð 62. Mahon, R. J., K. M. Olsen, K. A. Garsia, and S. R. Young. 2007. Resistance to the Bt toxin Cry2Ab in a strain of Helicoverpa armigera (Hu¨ bner) (Lepidoptera: Noctuidae) in Australia. J. Econ. Entomol. 100: 894 Ð902. Martinez, C., J. E. Ibarra, and P. Caballero. 2005. Association between serotype, cry gene content, and toxicity to Helicoverpa armigera larvae among Bacillus thuringiensis isolates native to Spain. J. Invertebr. Pathol. 90: 91Ð97. McGaughey, W. H., and M. E. Whalon. 1992. Managing insect resistance to Bacillus thuringiensis toxins. Science (Wash., D.C.) 258: 1451Ð1455.
1853
McKenzie, J. A., A. G. Parker, and J. L. Yen. 1992. Polygenic and single gene responses to selection for resistance to diazinon in Lucilia cuprina. Genetics 130: 613Ð 620. Olsen, K. M., J. C. Daly, H. E. Holt, and E. J. Finnegan. 2005. Season-long variation in expression of the cry1Ac gene and efÞcacy of Bacillus thuringiensis toxin in transgenic cotton against Helicoverpa armigera (Lepidoptera: Noctuidae). J. Econ. Entomol. 98: 1007Ð1017. Progue, M. G. 2004. A new synonym of Helicoverpa zea (Boddie) and differentiation of adult males of H. zea and H. armigera (Hubner) (Lepidoptera: Noctuidae: Heliothinae). Ann. Entomol. Soc. Am. 97: 1222Ð1226. Roush, R. T. 1998. Two-toxin strategies for management of insecticidal transgenic crops: can pyramiding succeed where pesticide mixtures have not? Phil. Trans. R. Soc. Lond. B 353: 1777Ð1786. Roush, R. T., and J. A. McKenzie. 1987. Ecological genetics of insecticide and acaricide resistance. Annu. Rev. Entomol. 32: 361Ð380. Sokal, R. R., and F. J. Rohlf. 1969. Biometry. Freeman and Co., San Francisco, CA. Stodola, T. J., and D. A. Andow. 2004. F2 screen variations and associated statistics. J. Econ. Entomol. 97: 1756 Ð1764. Stodola, T. J., D. A. Andow, A. R. Hyden, J. L. Hinton, J. J. Roark, L. L. Buschman, P. Porter, and G. B. Cronholm. 2006. Frequency of resistance to Bacillus thuringiensis toxin Cry1Ab in southern United States corn belt population of European corn borer (Lepidoptera: Crambidae). J. Econ. Entomol. 99: 502Ð507. Tabashnik, B. E. 1994. Evolution of resistance to Bacillus thuringiensis. Annu. Rev. Entomol. 39: 47Ð79. Tabashnik, B.E.T., J. Dennehy, and Y. Carrie`re. 2005. Delayed resistance to transgenic cotton in pink bollworm. Proc. Natl. Acad. Sci. U.S.A. 102: 15389 Ð15393. Teakle, R. E., and J. M. Jensen. 1985. Heliothis punctiger, pp. 312Ð322. In R. Singh and R. F. Moore [eds.], Handbook of insect rearing, vol. 2. Elsevier, Amsterdam, The Netherlands. Wang, J., A. Boets, J. Van Rie, and G. Ren. 2003. Characterization of cry1, cry2, and cry9 genes in Bacillus thuringiensi isolates from China. J. Invertebr. Pathol. 82: 63Ð71. Wenes, A.-L., D. Bourguet, D. A. Andow, C. Courtin, G. Carre, P. Lorme, L. Sanchez, and S. Augustin. 2006. Frequency and Þtness cost of resistance to Bacillus thuringiensis in Chrysomela tremulae (Coleoptera: Chrysomelidae). Heredity 97: 127Ð134. Wu, K. M., and Y. Y. Guo. 2005. The evolution of cotton pest management in China. Annu. Rev. Entomol. 50: 31Ð52. Wu, K. M., Y. Y. Guo, and G. Head. 2006. Resistance monitoring of Helicoverpa armigera (Lepidoptera: Noctuidiae) to Bt insecticidal protein during 2001Ð2004 in China. J. Econ. Entomol. 99: 893Ð 898. Zimowska, G. J., and A. M. Handler. 2006. Highly conserved piggyBack elements in noctuid species of Lepidoptera. Insect Biochem. Mol. Biol. 36: 421Ð 428. Received 14 May 2007; accepted 20 July 2007.