Effects of Gossypol on Fitness Costs Associated with ...

1 downloads 0 Views 156KB Size Report
Author(s): Yves CarriËre, Christa Ellers-Kirk, Robert Biggs, Dawn M. Higginson,. Timothy J. Dennehy, and Bruce E. Tabashnik. Source: Journal of Economic ...
Effects of Gossypol on Fitness Costs Associated with Resistance to Bt Cotton in Pink Bollworm Author(s): Yves Carrière, Christa Ellers-Kirk, Robert Biggs, Dawn M. Higginson, Timothy J. Dennehy, and Bruce E. Tabashnik Source: Journal of Economic Entomology, 97(5):1710-1718. 2004. Published By: Entomological Society of America DOI: http://dx.doi.org/10.1603/0022-0493-97.5.1710 URL: http://www.bioone.org/doi/full/10.1603/0022-0493-97.5.1710

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research.

INSECTICIDE RESISTANCE AND RESISTANCE MANAGEMENT

Effects of Gossypol on Fitness Costs Associated with Resistance to Bt Cotton in Pink Bollworm YVES CARRIE`RE, CHRISTA ELLERS-KIRK, ROBERT BIGGS, DAWN M. HIGGINSON, TIMOTHY J. DENNEHY, AND BRUCE E. TABASHNIK Department of Entomology, The University of Arizona, Tucson, AZ 85716

J. Econ. Entomol. 97(5): 1710Ð1718 (2004)

ABSTRACT Fitness costs associated with insect resistance to Bacillus thuringiensis (Bt) crops may help to delay or prevent the spread of resistance alleles, especially when refuges of non-Bt host plants are present. The potential for such delays increases as the magnitude and dominance of Þtness costs increase. Here, we examined the idea that plant secondary chemicals affect expression of Þtness costs associated with resistance to Bt cotton in Pectinophora gossypiella (Saunders). SpeciÞcally, we tested the hypotheses that gossypol affects the magnitude or dominance of Þtness costs, by measuring performance of three independent sets of pink bollworm populations fed artiÞcial diet with and without gossypol. Each set had an unselected susceptible population, a resistant population derived by selection from the susceptible population, and the F1 progeny of the susceptible and resistant populations. No individuals completed development on diets with gossypol in one set, suggesting that these individuals partially lost the ability to detoxify this chemical. In the other two sets, costs affecting survival did not support the hypotheses, but costs affecting pupal weight did. Adding gossypol to diet increased the magnitude and dominance of costs affecting pupal weight. In one of the two sets with survivors on diet with gossypol, costs affecting development time were less recessive when gossypol was present in diet. These results indicate that gossypol increased the magnitude and dominance of some Þtness costs. Better understanding of the effects of natural plant defenses on Þtness costs could improve our ability to design refuges for managing insect resistance to Bt crops. KEY WORDS Bacillus thuringiensis, Þtness costs, gossypol, Pectinophora gossypiella, transgenic crops

ALLELES CONFERRING RESISTANCE TO insecticides may have negative pleiotropic effects that lower Þtness in resistant individuals relative to susceptible individuals in the absence of insecticides (Carrie` re et al. 1994, 1995; Carrie` re and Roff 1995; McKenzie 1996). Resistance alleles are generally rare when novel insecticides are introduced, suggesting that such Þtness costs are common in targeted pests (McKenzie 1996). Resistance to Bacillus thuringiensis (Bt) seems to Þt this prediction. So far, only one pest species has evolved resistance to Bt in the Þeld (Tabashnik 1994, Tabashnik et al. 2003) and Þtness costs seem common (Tabashnik et al. 1994, Tabashnik 1998, Ferre´ and Van Rie 2002). Such Þtness costs can be substantial, affecting most life history traits, as well as mating success and sperm precedence (Groeters et al. 1994; Trisyono and Whalon 1997; Alyokhin and Ferro 1999; Carrie` re et al. 2001a, b; Higginson 2003; Janmaat and Myers 2003). Fitness costs may help to delay or prevent the spread of alleles conferring resistance to Bt crops when refuges of non-Bt host plants are present (Lenormand and Raymond 1998, Carrie` re and Tabashnik 2001). Insecticide resistance is often conferred by a few alleles with major effects (Carrie` re and Roff 1995, McKenzie 1996, Morin et al. 2003). When a

single gene confers resistance to a Bt crop, three resistance genotypes may be considered, RR, RS, and SS, where S is an allele for susceptibility and R for resistance. Fitness costs are recessive when negative pleiotropic effects of a resistance allele are only expressed in RR individuals. However, costs are nonrecessive when Þtness is lower in RS than in SS individuals (McKenzie 1990). Conditions favoring delays in the evolution of resistance include 1) large recessive Þtness costs and large refuges; 2) recessive Þtness costs, incomplete resistance, and some refuges; and 3) nonrecessive Þtness costs and some refuges (Carrie` re and Tabashnik 2001, Carrie` re et al. 2002). Nonrecessive Þtness costs may be particularly potent for delaying resistance to Bt crops. For example, a small Þtness cost of 1% in heterozygotes relative to susceptible homozygotes in non-Bt host plant refuges may delay or block the spread of a resistance allele under a broad range of resistance allele frequencies and refuge sizes (Fig. 1; Carrie` re and Tabashnik 2001). Our long-term goal is to determine whether refuge plants (including native vegetation) could be chosen, designed, or managed to make Þtness costs larger or less recessive in pests targeted by Bt crops (Carrie` re et al. 2001b, 2002, 2004, Tabashnik et al. 2003). We

0022-0493/04/1710Ð1718$04.00/0 䉷 2004 Entomological Society of America

October 2004

` ET AL.: FITNESS COSTS ASSOCIATED WITH RESISTANCE TO BT CROPS CARRIERE

1711

Fig. 1. Effect of dominance of Þtness cost on evolution of resistance to a Bt crop. As previously documented in the pink bollworm, resistance was recessive and incomplete resistance was present: survival of the RR, RS, and SS individuals on the Bt crop was 0.4, 0, and 0, respectively. With recessive Þtness costs, survival of the RR, RS, and SS individuals in refuges was 0.5, 1, and 1, respectively. Survival of RS individuals in refuges was 0.99 and 0.90 with Þtness cost of 1 and 10% in RS, respectively. Initial frequency of the resistance allele was 0.01 and refuge size 20%. See Carrie` re and Tabashnik (2001) and Carrie` re et al. (2002) for details on simulation model.

previously found large recessive Þtness costs associated with resistance to Bt cotton in pink bollworm, Pectinophora gossypiella (Saunders) (Carrie` re et al., 2001a, b). Moreover, pink bollworm resistance to Bt cotton is functionally recessive and linked with deletions in a gene encoding a cadherin protein that binds Cry1Ac, the Bt toxin in Bt cotton (Tabashnik et al. 2000, 2002; Liu et al. 2001; Morin et al. 2003). Here, we examined the hypothesis that Þtness costs associated with resistance to Bt crops are affected by plant defenses. SpeciÞcally, we tested the hypotheses that increasing the concentration of gossypol, a cotton phytochemical toxic to many herbivores (Shaver and Parrott, 1970, Meisner et al. 1976, Hedin et al. 1991), increases the magnitude and dominance of Þtness costs in pink bollworm. To test these hypotheses, we compared the effects of gossypol in artiÞcial diet on performance of three types of pink bollworm larvae (susceptible, Bt-resistant, and their F1 progeny) from three independent origins. Materials and Methods Pink Bollworm Strains and Crosses. From each of three independent sets of larvae (APHIS, MOV, and SAF) derived from Arizona Þeld populations, we tested three types of larvae: an unselected parent strain that was predominantly susceptible (APHIS-S, MOV97, and SAF97), a resistant strain derived by laboratory selection with Cry1Ac from the parent strain (APHIS-98R, MOV-97R100, and SAF-97R), and their F1 progeny (APHIS-S ⫻ APHIS-98R, MOV97 ⫻ MOV-97R100, and SAF97 ⫻ SAF-97R). For convenience, each of the nine aforementioned groups of larvae is referred to as a population. Larvae were reared in the laboratory on wheat germ diet (Adkin-

son et al. 1960). Susceptible and resistant populations were maintained at 200 to ⬎1000 adults per generation. APHIS-S had been reared in the laboratory for ⬎20 yr without exposure to insecticide and harbored few or no alleles conferring resistance to Bt (Liu et al. 2001, Morin et al. 2003; see below). The susceptible strains from Mohave Valley (MOV97) and Safford (SAF97) had not been exposed to Bt toxins in the laboratory, but were derived in 1997 from Þeld populations harboring some resistance alleles (Tabashnik et al. 2000, Carrie` re et al. 2001a, Morin et al. 2003) and still had some resistance alleles at the time of this study (see below). APHIS-98R was created by selecting a subset of APHIS-S, Þrst with Bt cotton leaf powder, and then with Cry1Ac in diet (Liu et al. 2001). SAF97-R was created by feeding ⬇5800 neonates of the F33 generation of SAF97 with diet containing 10 ␮g of Cry1Ac per milliliter of diet. This yielded 288 survivors. Thereafter, SAF97-R was selected with 10 ␮g of Cry1Ac per milliliter of diet every three to six generations. MOV97-R10 (not tested here) was started by selecting a subset of the F10 generation of MOV97 with10 ␮g Cry1Ac per milliliter of diet, with selection at that concentration continuing in every other generation (Carrie` re et al. 2001a). MOV97-R100 was started by feeding neonates of the F28 generation of MOV97-R10 with 100 ␮g of Cry1Ac per milliliter of diet. Thereafter, MOV97-R100 was selected with 100 ␮g of Cry1Ac per milliliter of diet every three to six generations. At the time of this study, generations of rearing were 43 for MOV97, 47 for SAF97, 36 for APHIS-98R, 10 for MOV97-R100, and 13 for SAF97-R (counting generations after splitting in resistant populations). Individual from APHIS-98R, MOV97-R10, and SAF97-R can

1712

JOURNAL OF ECONOMIC ENTOMOLOGY

survive on Bt cotton (Tabashnik et al. 2003; unpublished data). To obtain F1 progeny, we performed reciprocal mass crosses between each pair of related susceptible and resistant strains. In each case, 200 male pupae from the susceptible strain were pooled with 200 female pupae from the resistant strain and vice versa. Frequency of Resistance Alleles. We used bioassays of random samples of the experimental neonates to estimate the frequency of R alleles for each of the six susceptible and resistant populations. For each of these populations, 40 larvae were fed individually on diet without Cry1Ac and 40 larvae were fed individually on diet with a concentration of Cry1Ac (10 ␮g/ml diet) that allows survival only of RR individuals (Patin et al. 1999; Tabashnik et al. 2000, 2002; Morin et al. 2003). Resistance allele frequency was estimated as the square root of adjusted survival at 10 ␮g of Cry1Ac per ml diet (survival at 10 ␮g/ml divided by survival at 0 ␮g/ml). As expected, estimated resistance allele frequencies were low for unselected populations (0 for APHIS-S, 0.16 for MOV97, 0.24 for SAF97) and high for resistant populations (1.0 for APHIS-98R, 1.0 for MOV97-R100, and 0.96 for SAF97-R). Artificial Diet and Bioassays. Gossypol (95% in acetic acid crystal) was incorporated into wheat germ diet (Adkinson et al. 1960) at 0, 0.1, or 0.2% (grams of gossypol/100 g diet; Shaver and Parrott 1970). For APHIS and MOV bioassays, we used two batches of 720 g of each diet. For SAF bioassays, we used 1278 g of each diet. Neonates (⬍24 h old) from each population were put in trays (C-D International, Pitman NJ), each with 16 wells (15 mm in depth, 3 ml in volume). Each well was Þlled with ⬇4 g of diet and received one neonate. After all wells from a plate had received a neonate, the plate was sealed with a transparent cover with ventilation holes (Liu et al. 2001). Each plate was put in a plastic box with ventilation holes. The boxes were stacked in groups of three, each group comprising neonates from the same origin (either APHIS, MOV, or SAF) and the three population types (susceptible, resistant, and F1), and larvae feeding on the same diet type (either 0, 0.1, or 0.2% gossypol). These groups were randomly arranged in a growth chamber at 29⬚C, 65% RH, and a photoperiod of 14:10 (L:D) h. For bioassays with APHIS and MOV, Þve plates were used for each combination of population and diet type (i.e., Þve plates ⫻ 16 wells ⫽ 80 larvae per population and diet type). For SAF, the number of plates used was seven for control diet and six one-half for diets with gossypol (n ⫽ 112 and 104 neonates, respectively, per population and diet type). Statistical Analyses. All analyses were performed with JMP (JMP Statistics and Graphics Guide 2001). For MOV and SAF, logistic regression for binomial counts (Ramsey and Schafer 2002) was Þrst used to evaluate variation in survival as a function of gossypol concentration and population. The odds of survival in a plate was the response variable, whereas gossypol concentration, population, and the interaction between gossypol concentration and population were

Vol. 97, no. 5

the explanatory variables. Additionally, pairwise FisherÕs exact tests were used to evaluate differences in survival among populations at each diet concentration. Multiple regression analysis was used to assess variation among populations in developmental time (from hatching of neonates to adult eclosion) and pupal weight (only considering pupae producing adults). The explanatory variables included in the initial models were gossypol concentration, population, sex, and all two-way interactions and the three-way interactions involving these explanatory variables. The response variables were transformed when necessary to improve Þt to assumptions of normality and homogeneity of variance (Ramsey and Schafer 2002). A signiÞcant interaction between gossypol concentration and population in the above-mentioned analyses would indicate that increasing the concentration of gossypol changed the magnitude or dominance of Þtness costs. However, due to the relatively small effect of gossypol on expression of Þtness costs, the power of the test of these interactions was low compared with the power of the test of the main effects (see Results). To improve our ability to detect the small but nonetheless ecologically relevant changes in expression of Þtness costs (Fig. 1; Carrie` re and Tabashnik, 2001), least squares means contrasts were used to further assess differences in development time and pupal weight among the populations at each gossypol concentration. Large Þtness costs were detected previously when larvae ate non-Bt cotton but not when they ate wheat germ diet without Cry1Ac or gossypol (Tabashnik et al. 2000; Carrie` re et al. 2001a, b). Thus, two-tailed and one-tailed contrasts were used to compare the life history traits among populations fed diets without and with gossypol, respectively. Results APHIS populations produced adults when larvae ate diet without gossypol, but not when larvae ate diet with gossypol. On diet with gossypol, APHIS larval development seemed normal through pupation, but all pupae died. Many APHIS pupae were deformed and incompletely schlerotized, suggesting that disruption of the molting process resulted in pupal death. In contrast, MOV and SAF populations produced adults when larvae ate diet with gossypol (see below). The duration of laboratory rearing on diet was only 4 Ð5 yr for MOV and SAF populations, compared with ⬎20 yr for APHIS populations. Thus, it is likely that the APHIS populations lost the ability to detoxify gossypol incorporated in artiÞcial diet. APHIS Populations. Because no individuals from the APHIS populations survived on diet with gossypol, the data on developmental time and pupal weight on diets with gossypol were not analyzed further. Proportion survival on diet without gossypol was 0.30 for APHIS-S, 0.16 for F1, and 0.05 for APHIS-98R. Survival was lower in APHIS-98R than in APHIS-S (FisherÕs exact test, P ⬍ 0.0001), indicating a Þtness cost in

October 2004

` ET AL.: FITNESS COSTS ASSOCIATED WITH RESISTANCE TO BT CROPS CARRIERE

Table 1. Survival of the populations fed on artificial diet with different concentrations of gossypol (n ⴝ 80 –112 for each survival value)

Population MOV97 F1 (MOV97 ⫻ MOV97-R100 MOV97- R100 SAF97 F1 (SAF97 ⫻ SAF97-R) SAF97-R

P survivala Gossypol concn (g/100 g diet) 0

0.1

0.2

0.40a 0.40a 0.40a 0.29a 0.30a 0.47b

0.21a 0.35a 0.26a 0.06a 0.22b 0.34b

0.07a 0.07a 0.17a 0.12a 0.24b 0.19a

a Related strains sharing the same letter within a diet concentration had similar proportions of survival (FisherÕs exact tests, P ⬎ 0.05).

APHIS-98R. Survival was also lower in F1 than APHIS-S (FisherÕs exact test, P ⫽ 0.060), indicating that this Þtness cost was not completely recessive. No signiÞcant difference in pupal weight or development time occurred among the APHIS populations on diet without gossypol (data not shown; P ⬎ 0.05). MOV Populations. As gossypol concentration increased, survival decreased in the MOV populations (Table 1; ␹2 ⫽ 31.1, P ⬍ 0.0001). However, mean survival across diets was similar for the three populations (␹2 ⫽ 2.034, P ⫽ 0.36) and no interaction occurred between gossypol concentration and population (␹2 ⫽ 2.00, P ⫽ 0.37). No among-population difference occurred on each diet type (Table 1). Survival was especially low in MOV97 and F1 at a concentration of 0.2% gossypol (Table 1). Only six individuals survived in each case and no signiÞcant differences in developmental time or pupal weight were detected (P ⬎ 0.05). Because sample size was small at 0.2% gossypol, we focus on analyses of developmental time and pupal weight on diets with 0 and 0.1% gossypol. The Þnal multiple regression model for the analysis of developmental time (log transformed) contained gossypol concentration, population, and the interaction between these two factors. Developmental time was greater on a diet with 0.1% gossypol than without gossypol (Fig. 2; F ⫽ 205.8; df ⫽ 1, 155; P ⬍ 0.0001: Power ⫽ 1). Mean developmental time across diets differed among the populations (F ⫽ 5.66; df ⫽ 1, 155; P ⫽ 0.0043: Power ⫽ 0.86), with least squares means (backtransformed) of 31.4, 31.9, and 33.3 d for MOV97, F1, and MOV97-R100, respectively. Evidence was moderate for the presence of a signiÞcant interaction between gossypol concentration and population (F ⫽ 1.85; df ⫽ 1, 155; P ⫽ 0.16: Power ⫽ 0.38). MOV97-R100 developed slower than MOV97 with or without gossypol, indicating a Þtness cost on both diets (Fig. 2). Developmental time was signiÞcantly greater for F1 than MOV97 with gossypol, but it did not differ between F1 and MOV97 without gossypol (Fig. 2). With 0.1% gossypol, the least square means (backtransformed) were 33.9 and 36.3 d for MOV97 and F1, respectively, suggesting a cost of 7.1%. Thus, there was evidence that the Þtness cost affecting development time was less recessive with gossypol.

1713

The Þnal model for pupal weight (log transformed) contained gossypol concentration, population, sex, the interaction between gossypol concentration and population. Gossypol reduced pupal weight (Fig. 2; F ⫽ 244.6; df ⫽ 1, 152; P ⬍ 0.0001: Power ⫽ 1). Mean pupal weight across the diets did not differ among the populations (F ⫽ 2.06; df ⫽ 1, 152; P ⫽ 0.13: Power ⫽ 0.41), and no interaction occurred between gossypol concentration and population (F ⫽ 1.15; df ⫽ 1, 152; P ⫽ 0.32: Power ⫽ 0.25). Females were heavier than males (F ⫽ 74.22; df ⫽ 1, 152; P ⬍ 0.0001: Power ⫽ 1), although none of the interactions involving sex were signiÞcant (P ⬎ 0.087). Contrasts on least squares means of pupal weight did not reveal among-population differences without gossypol (Fig. 2). However, on diet with 0.1% gossypol, pupal weight was lower in F1 and MOV97-R100 than in MOV97 (Fig. 2.). With 0.1% gossypol, the least square means (backtransformed) were 0.0172, 0.0155, and 0.0157 g for MOV97, F1, and MOV97-R100, respectively. This suggests that gossypol was associated with a nonrecessive Þtness cost reducing pupal weight by 8.7% in F1, and a cost of similar magnitude in MOV97-R100. SAF Populations. As gossypol concentration increased, survival in the SAF populations decreased (Table 1; ␹2 ⫽ 15.34, P ⫽ 0.0001). Mean survival across diets was higher in SAF97-R than SAF97 (␹2 ⫽ 5.86, P ⬍ 0.0001) and higher in F1 than SAF97 (␹2 ⫽ 5.86, P ⫽ 0.015). The interaction between gossypol concentration and population was not signiÞcant (␹2 ⫽ 3.64, P ⫽ 0.16). Nonetheless, SAF97-R survived better than the other populations without gossypol, SAF97 survived less well than the other populations on diet with 0.1% gossypol, and the F1 population survived better than the other two on diet with 0.2% gossypol. The Þnal model for developmental time contained gossypol concentration, population, sex, and the interaction between gossypol concentration and population. Developmental time increased on diets with gossypol (Fig. 3; df ⫽ 2, 224; F ⫽ 42.1; P ⬍ 0.0001: Power ⫽ 1). Mean developmental time across the diets differed among populations (df ⫽ 2, 224; F ⫽ 10.8; P ⬍ 0.0001: Power ⫽ 0.99), with the least squares means for SAF97, F1, and SAF97-R being 29.8, 31.3, and 32.9 d, respectively. No interaction between gossypol concentration and population was evident (F ⫽ 0.19; df ⫽ 4, 224; P ⫽ 0.94: Power ⫽ 0.09). Developmental time was longer in females than males (F ⫽ 97.2; df ⫽ 1, 224; P ⫽ 0.0048: Power ⫽ 0.81), although none of the interactions involving sex was signiÞcant (P ⬎ 0.52). The least squares means contrasts suggested that Þtness costs affected development time in SAF97-R on all diets (Fig. 3). Developmental time was also longer in F1 than SAF97 on diet without gossypol, but not on diets with gossypol (Fig. 3). So, in this case, gossypol did not make Þtness costs larger or less recessive. The Þnal model for pupal weight (log transformed) included gossypol concentration, population, sex, and the interaction between gossypol concentration and population. As gossypol concentration increased, pupal weight decreased in the SAF populations (Fig. 3;

1714

JOURNAL OF ECONOMIC ENTOMOLOGY

Vol. 97, no. 5

Fig. 2. Effect of gossypol concentration incorporated into artiÞcial diet on developmental time (log d) and pupal weight (log (g ⫻ 1000)) in the MOV strains. The asterisks (*P ⬍ 0.05, **P ⬍ 0.01) indicate the outcome of the statistical comparisons between the SS and other genotypes at each gossypol concentration. At the concentration of 0.1%, the two stars apply to the comparison between the SS and RR individuals.

F ⫽ 127.2; df ⫽ 2, 224; P ⬍ 0.0001: Power ⫽ 1). Mean pupal weight across the diets differed among the populations (F ⫽ 7.8; df ⫽ 2, 224; P ⫽ 0.0005: Power ⫽ 0.95), with least squares means (back transformed) of 0.018, 0.018, and 0.016 g for SAF97, F1, and SAF97-R10, respectively. Evidence was moderate for the presence of a signiÞcant interaction between gossypol concentration and population (F ⫽ 1.61; df ⫽ 4, 224; P ⫽ 0.17: Power ⫽ 0.49). Females were larger than males (F ⫽ 30.0; df ⫽ 1, 224; P ⬍ 0.001: Power ⫽ 1), although none of the interactions involving sex were signiÞcant (P ⬎ 0.11). Although the least squares means of pupal weight were similar without gossypol, increasing gossypol concentration generated Þtness costs in SAF97-R but not in F1 (Fig. 3). With 0.1 and 0.2% gossypol, the least square means (backtransformed) were, respectively, 0.0175 and 0.0145 g for SAF97 and 0.0145 and 0.0127 g for SAF97-R. Therefore, there was evidence that Þtness costs affecting pupal weight increased in SAF97-R by 17.1 and 12.4% with 0.1 and

0.2% gossypol, respectively. However, gossypol did not make such costs less recessive. Discussion Low frequencies of resistance alleles are expected when Bt crops are introduced to control pests not previously targeted by Bt sprays (Tabashnik et al. 2003). In such cases, heterozygous individuals would carry most resistance alleles. When resistance to a Bt crop is recessive or nearly recessive, the heritability of resistance is low if most resistant individuals surviving in Bt crop Þelds mate with susceptible insects from refuges (Carrie` re et al. 2004, Sisterson et al. 2004, Tabashnik et al. 2004). Under such conditions, evolution of resistance is slow and even small nonrecessive Þtness costs may considerably delay or prevent the spread of resistance alleles (Carrie` re and Tabashnik 2001, Carrie` re et al. 2002). Here, we manipulated artiÞcial diet fed to three independent sets of suscepti-

October 2004

` ET AL.: FITNESS COSTS ASSOCIATED WITH RESISTANCE TO BT CROPS CARRIERE

1715

Fig. 3. Effect of gossypol concentration incorporated in artiÞcial diet on developmental time (d) and pupal weight (log (g ⫻ 1000)) in the SAF strains. The asterisks (*P ⬍ 0.05, **P ⬍ 0.01) indicate the outcome of the statistical comparisons between the SS and other genotypes at each gossypol concentration.

ble, Bt-resistant, and F1 pink bollworm populations to determine whether gossypol, an important cotton phytochemical, could increase the size and dominance of Þtness costs. Survival of MOV and SAF populations on diet containing gossypol did not support this hypothesis (Table 1), nor did developmental time of SAF populations (Fig. 3). However, the across-diet changes in developmental time for MOV populations (Fig. 2), and in pupal weight for MOV and SAF populations (Figs. 2 and 3) supported this hypothesis. Thus, our results indicate that in some, but not all cases, gossypol increased the magnitude or dominance of Þtness costs associated with resistance to Bt in pink bollworm. Knowledge of the mutations causing resistance may help to elucidate the mechanisms leading to plasticity in Þtness costs. Resistance to Bt toxin Cry1Ac is linked to mutations in a cadherin-encoding gene in pink bollworm (Morin et al. 2003) and in Heliothis virescens (F.) (Gahan et al. 2001). Cadherin proteins have many functions and are widely distributed within organisms and across taxa (Nollet et al. 2000). An important

subfamily of cadherins contributes in maintaining cellcell adhesion in mammalian epithelia subjected to mechanical stress (Takeichi 1990, Angst et al. 2001). Although it is suspected that cadherin proteins could play a similar role in insect midgut epithelia, the function of cadherins in insects remains uncertain (Candas et al., 2002, Dorsch et al., 2002, but see Higginson 2003). In Helicoverpa zea (Boddie), gossypol enters the hemolymph where it competes for speciÞc binding sites in the fat bodies with the polyaromatic hydrocarbon 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (Muehleisen et al. 1989). TCDD also competes for binding sites with juvenile hormone 1 and methoprene (Muehleisen et al. 1989). Thus, a possible interaction between gossypol and juvenile hormone, in conjunction with a reduction in the ability to detoxify gossypol, could explain the abnormal pupal formation in the APHIS individuals fed on a diet with gossypol. We suggest that gossypol may penetrate the hemolymph and affect vital functions, and that the cadherin mutation causing Bt resistance in pink bollworm may

1716

JOURNAL OF ECONOMIC ENTOMOLOGY

Vol. 97, no. 5

Fig. 4. Frequency of major Bt resistance alleles as a function of time reared on artiÞcial diet (generations without toxin) in the MOV97 and SAF97 strains. When frequency estimates were obtained using larvae from more than one generation (e.g., bioassays conducted with larvae from generations zero to seven; Tabashnik et al. 2000), median generation is represented (e.g., 3.5). Data from Tabashnik et al. (2000), Carrie` re et al. (2001a, b), and Higginson (2003).

increase gossypol penetration, thereby magnifying the negative effects of gossypol (Carrie` re et al. 2004). To test this idea, we will need to analyze larval hemolymph (Parrott et al. 1987) to test whether the concentration of gossypol is larger in RS and RR individuals than in SS individuals. This will help assess the alternative possibility that changes in Þtness costs occurred because gossypol directly reduced activity of the cadherin proteins, which in turn would be essential to maintaining efÞcient digestive processes. Finally, although gossypol has direct toxic effects on a wide range of insect herbivores (see references therein), it also can inhibit larval feeding (Meisner et al. 1976, Parrott et al. 1987). Thus, differential feeding of the genotypes could have caused some of the variation in the across-diet changes in the Þtness components. We tested the effect of concentrations of 0.1 and 0.2% gossypol in artiÞcial diet (fresh weight), which, respectively, correspond to concentrations of ⬇0.6 and 1.2% on a dry weight basis. Gossypol concentration in cottonseed varies between 1.0 and 1.6% (dry weight) in Pima and Upland cotton grown in the U.S. Southwest (Lukefahr and Houghtaling 1969, Hron et al. 1999). Pink bollworm larvae feed on cottonseeds after entering bolls (Ingram 1994), suggesting that the gossypol concentrations used in this experiment were realistic. Nevertheless, larvae from the APHIS populations did not complete development on diets with gossypol, whereas they develop well on non-Bt and Bt-cotton (Carrie` re et al. 2001b; Liu et al. 1999, 2001; Y.C., unpublished data). The fact that the APHIS populations survive on cotton indicates that their gossypol detoxiÞcation ability has not been completely lost. Moreover, it suggests that larvae feeding on cotton are exposed to lower concentrations of gossypol than those used in this experiment, or that gossypol incor-

porated in artiÞcial diet acts differently than gossypol in cotton tissues. Further experiments in which larvae of different Bt resistance genotypes are fed different cotton varieties will be useful to better understand how natural plant defenses affect expression of Þtness costs in pink bollworm. Unexpectedly, Þtness costs affected survival in the APHIS populations on diet without gossypol. Moreover, Þtness costs also affected development time in the MOV and SAF populations on diet without gossypol (Figs. 2 and 3). Such costs should have favored a temporal decline in the frequency of the resistance alleles in the heterogeneous MOV97 and SAF97 laboratory strains. Yet, a decline in resistance was observed in MOV97 but was not obvious in SAF97 (Fig. 4). With nonrecessive costs affecting developmental time in the SAF populations but recessive costs in the MOV populations on diet without gossypol (Figs. 2 and 3), one would predict a faster decline in resistance frequency in SAF97 than in MOV97. Thus, the absence of a clear decline in resistance in SAF97 is puzzling. Perhaps one or more Þtness components other than those measured here are enhanced rather than diminished in resistant individuals in SAF97. Effects of gossypol on pupal weight were similar for MOV and SAF populations, but effects of gossypol on developmental time differed between MOV and SAF (Figs. 2 and 3). This difference might be related to differences in the resistance alleles in the two sets of populations. MOV and SAF share one resistance allele (R1), whereas another (R2) occurs in SAF but not MOV, and a third (R3) occurs in MOV but not SAF (Morin et al. 2003). Alternatively, the difference between the MOV and SAF populations could be due to different genetic backgrounds, which could be present either because the populations originated from different geographic locations or drift occurred

October 2004

` ET AL.: FITNESS COSTS ASSOCIATED WITH RESISTANCE TO BT CROPS CARRIERE

after their establishment in the laboratory. For example, the generally low survival in SAF97 (Table 1) could be due to drift resulting in inbreeding depression. Ongoing work based on recent development of genetic markers (Morin et al. 2003) will improve our ability to discriminate between these possibilities. The present results suggest that plant phytochemicals can affect the magnitude and dominance of Þtness costs associated with resistance to Bt crops. Although we focused here on plasticity of Þtness costs (sensu Schlichting and Pigliucci 1998) in the pink bollworm, the production of strains of other pests resistant to Bt crops would allow a more general evaluation of the hypothesis that Þtness costs are inßuenced by natural plant defenses. A better understanding of the effects of plant defenses on Þtness costs could improve our ability to manage or produce refuges to slow the evolution of insect resistance to Bt crops. Acknowledgments This study was supported by grant 99-35302Ð 8300 and 01-35302Ð 09976 from the USDAÐNRI program, and grant 2003-04371 from USDA Biotechnology Risk Assessment Research.

References Cited Adkinson, P. L., E. S. Vaderzant, D. L. Bull, and W. E. Allision. 1960. A wheat germ diet medium for rearing the pink bollworm. J. Econ. Entomol. 53: 759 Ð762. Alyokhin, A. V., and D. N. Ferro. 1999. Relative Þtness of Colorado potato beetle (Coleoptera: Chrysomelidae) resistant and susceptible to the Bacillus thuringiensis Cry3A toxin. J. Econ. Entomol. 92: 510 Ð515. Angst, B. D., C. Marcozzi, and A. I. Magee. 2001. The cadherin superfamily: diversity in form and function. J. Cell Sci. 114: 629 Ð 641. Candas, M., B. R. Francis, N. B. Griko, E. G. Midboe, and L. A. Bulla Jr. 2002. Proteolitic cleavage of the developmentally important cadherin BT-R1 in the midgut epithelium of Manduca sexta. Biochemistry 41: 13717Ð13724. Carrie`re, Y., and D. A. Roff. 1995. Change in genetic architecture resulting from the evolution of insecticide resistance: a theoretical and empirical analysis. Heredity 74: 357Ð368. Carrie`re, Y., and B. E. Tabashnik. 2001. Reversing insect adaptation to transgenic insecticidal crops. Proc. R. Soc. Lond. B. 268: 1475Ð1480. Carrie`re, Y., J.-P. Deland, D. A. Roff, and C. Vincent. 1994. Life history costs associated with the evolution of insecticide resistance. Proc. R. Soc. Lond. B. 58: 35Ð 45. Carrie`re, Y., D. A. Roff, and J.-P. Deland. 1995. The joint evolution of diapause and insecticide resistance: a test of an optimality model. Ecology 76: 1497Ð1505. Carrie`re, Y., C. Ellers-Kirk, A. L. Patin, M. A. Sims, S. Meyer, Y.-B. Liu, T. J. Dennehy, and B. E. Tabashnik. 2001a. Overwintering costs associated with resistance to transgenic cotton in the pink bollworm. J. Econ. Entomol. 94: 935Ð941. Carrie`re, Y., C. Ellers-Kirk, Y.-B. Liu, M. A. Sims, A. L. Patin, T. J. Dennehy, and B. E. Tabashnik. 2001b. Fitness costs and maternal effects associated with resistance to transgenic cotton in the pink bollworm. J. Econ. Entomol. 94: 1571Ð1576.

1717

Carrie`re, Y., T. J. Dennehy, C. Ellers-Kirk, D. Holley, Y.-B. Liu, M. A. Sims, and B. E. Tabashnik. 2002. Fitness costs, incomplete resistance, and management of resistance to Bt crops, pp. 82Ð91. In R. J. Akhurst, C. E. Beard, and P. A. Huges [eds.]. Proceedings of the 4th PaciÞc Rim Conference: Biotechnology of Bacillus thuringiensis and its Environmental Impact. Nov. 11Ð15, 2001. CSIRO, Canberra, Australia. Carrie`re, Y., M. Sisterson, and B. E. Tabashnik. 2004. Resistance management for sustainable use of Bacillus thuringiensis crops, pp. 65Ð95. In A. R. Horowitz and I. Ishaaya [eds.]. Insect pest management: Þeld and protected crops. Springer, New York. Dorsch, J. A., C. Candas, N. B. Griko, W.S.A. Maaty, E. G. Midboe, R. K. Vadlamudi, and L. A. Bulla Jr. 2002. Cry1A toxins of Bacillus thuringiensis bind speciÞcally to a region adjacent to membrane-proximal extracellular domain of BT-R1 in Manduca sexta: involvement of a cadherin in the entomopathogenecity of Bacillus thuringiensis. Insect Biochem. Mol. Biol. 32: 1025Ð1036. Ferre´, J., and J. Van Rie. 2002. Biochemistry and genetics of insect resistance to Bacillus thuringiensis. Annu. Rev. Entomol. 47: 501Ð533. Gahan, L. J., F. Gould, and D. G. Heckel. 2001. IdentiÞcation of a gene associated with Bt resistance in Heliothis virescens. Science (Wash. DC) 293: 857Ð 860. Groeters, F. R., B. E. Tabashnik, N. Finson, and M. W. Johnson. 1994. Fitness costs of resistance to Bacillus thuringiensis in the diamondback moth (Plutella xylostella). Evolution 48: 197Ð2001. Hedin, P. A., W. L. Parrott, and J. N. Jenkins. 1991. Effects of cotton plant allelochemicals and nutrient on behavior and development of tobacco budworm. J. Chem. Ecol. 17: 1107Ð1121. Higginson, D. M. 2003. Male reproductive costs of Bacillus thuringiensis resistance in the pink bollworm (Lepidoptera: Gelechiidae). M.S. thesis, Department of Entomology, The University of Arizona, Tucson. Hron, R. J., H. L. Kim, M. C. Calhoun, and G. S. Fisher. 1999. Determination of (⫹)-, (⫺)-, and total gossypol in cottonseed by high-performance liquid chromatography. J. Am. Oil Chem. Soc. 76: 1351Ð1355. Ingram, W. R. 1994. Pectinophora (Lepidoptera: Gelechiidae), pp. 107Ð149. In G. A. Mattews and J. P. Tunstall [eds.], Insect pests of cotton. CAB, Wallingford, United Kingdom. Janmaat, A. F., and J. H. Myers. 2003. Rapid evolution and the cost of resistance to Bacillus thuringiensis in greenhouse populations of cabbage loopers, Trichoplusia ni. Proc. R. Soc. Lond. B. 270: 2263Ð2270. JMP Statistics and Graphics Guide. 2001. SAS Institute, Cary, NC. Lenormand, T., and M. Raymond. 1998. Resistance management: the stable zone strategy. Proc. R. Soc. Lond. B. 265: 1985Ð1990. Liu, Y.-B., B. E. Tabashnik, T. J. Dennehy, A. L. Patin, and A. C. Bartlett. 1999. Development time and resistance to Bt crops. Nature (Lond.) 400: 519. Liu, Y.-B., B. E. Tabashnik, T. J. Dennehy, A. L. Patin, M. A. Sims, S. K. Meyer, and Y. Carrie`re. 2001. Effects of Bt cotton and Cry1Ac toxin on survival and development of pink bollworm (Lepidoptera: Gelechiidae). J. Econ. Entomol. 94: 1237Ð1242. Lukefahr, M. J., and J. E. Houghtaling. 1969. Resistance of cotton strains with high gossypol content to Heliothis spp. J. Econ. Entomol. 62: 588 Ð591.

1718

JOURNAL OF ECONOMIC ENTOMOLOGY

McKenzie, J. A. 1990. Selection at a dieldrin resistance locus in overwintering populations of Lucilia cuprina (Wiedemann). Aust. J. Zool. 38: 493Ð501. McKenzie, J. A. 1996. Ecological and evolutionary aspects of insecticide resistance. Academic, Austin, TX. Meisner, J., K.R.S. Ascher, and M. Zur. 1976. Phagodeterrency induced by pure gossypol and leaf extracts of cotton strains with high gossypol content in the larva of Spodoptera littoralis. J. Econ. Entomol. 70: 149 Ð150. Morin, S., R. W. Biggs, M. S. Sisterson, L. Shriver, C. EllersKirk, D. Higginson, D. Holley, L. J. Gahan, D. G. Heckel, Y. Carrie`re, et al. 2003. Three cadherin alleles associated with resistance to Bacillus thuringiensis in pink bollworm. Proc. Natl. Acad. Sci. U.S.A. 100: 5004 Ð5009. Muehleisen, D. P., F. W. Plapp Jr., J. H. Benedict, and F. A. Carino. 1989. High afÞnity TCDD binding to fat body cytosolic proteins of the bollworm, Heliothis zea. Pestic. Biochem. Physiol. 35: 50 Ð57. Nollet, F., P. Kools, and F. van Roy. 2000. Phylogenetic analysis of the cadherin superfamily allows identiÞcation of six major subfamilies besides several solitary members. J. Mol. Biol. 299: 551Ð572. Patin, A. L., T. J. Dennehy, M. A. Sims, B. E. Tabashnik, Y.-B. Liu, L. Antilla, D. Gouge, T. J. Henneberry, and R. Staten. 1999. Status of pink bollworm susceptibility to Bt in Arizona. Proc. Belt. Cotton Conf. 2: 1025Ð1035. Parrott, W. L., P. A. Hedin, J. N. Jenkins, and J. C. McCarty Jr. 1987. Feeding and recovery of gossypol and tannin from tobacco budworm larvae. Southwest. Entomol. 12: 197Ð204. Ramsey, F. L., and D. W. Schafer. 2002. The statistical sleuth: a course in methods of data analysis. Duxburry Press, New York. Sisterson, M. S., L. Antilla, Y. Carrie`re, C. Ellers-Kirk, and B. E. Tabashnik. 2004. Effect of insect population size on evolution of resistance to transgenic crops. J. Econ. Entomol. 97: 1413Ð1424. Shaver, T. N., and W. L. Parrott. 1970. Relationship of larval age to toxicity of Gossypol to bollworms, tobacco bollworms, and pink bollworm. J. Econ. Entomol. 63: 1802Ð 1804.

Vol. 97, no. 5

Schlichting, C. D., and M. Pigliucci. 1998. Phenotypic evolution: a reaction norm perspective. Sinauer, New York. Tabashnik, B. E. 1994. Evolution of resistance to Bacillus thuringiensis. Annu. Rev. Entomol. 39: 47Ð79. Tabashnik, B. E. 1998. Transgenic crops for the paciÞc basin: prospects and problems, pp. 161Ð168. In M. P. Zalucki, R.A.I. Drew, and G. G. White [eds.], Proceedings of the 6th Australian Applied Entomological Research Conference, vol. 1. Brisbane, Australia. Tabashnik, B. E., F. R. Groeters, N. Finson, and M. W. Johnson. 1994. Instability of resistance to Bacillus thuringiensis. Bio. Sci. Technol. 4: 419 Ð 426. Tabashnik, B. E., A. L. Patin, T. J. Dennehy, Y.-B. Liu, Y. Carrie`re, M. A. Sims, and L. Antilla. 2000. Frequency of resistance to Bacillus thuringiensis in Þeld populations of pink bollworm. Proc. Natl. Acad. Sci. U.S.A. 97: 12980 Ð 12984. Tabashnik, B. E., Y.-B. Liu, T. J. Dennehy, M. A. Sims, M. Sisterson, R. Biggs, and Y. Carrie`re. 2002. Inheritance of resistance to Bt toxin Cry1Ac in a Þeld-derived strain of pink bollworm (Lepidoptera: Gelechiidae). J. Econ. Entomol. 95: 1018 Ð1026. Tabashnik, B. E., Y. Carrie`re, T. J. Dennehy, S. Morin, M. Sisterson, R. T. Roush, A. M. Shelton, and J.-Z. Zhao. 2003. Insect resistance to transgenic Bt crops: lessons from the laboratory and Þeld. J. Econ. Entomol. 96: 1031Ð 1038. Tabashnik, B. E., F. Gould, and Y. Carrie`re. 2004. Delaying evolution of insect resistance to transgenic crops by decreasing dominance and heritability. J. Evol. Biol. 17: 904 Ð912. Takeichi, M. 1990. Cadherins: a molecular family important in selective cell-cell adhesion. Annu. Rev. Biochem. 59: 237Ð252. Trisyono, A., and M. E. Whalon. 1997. Fitness costs of resistance to Bacillus thuringiensis in Colorado potato beetle (Coleoptera: Chrysomelidae). J. Econ. Entomol. 90: 267Ð271. Received 24 November 2003; accepted 27 June 2004.