Abstract

Laboratory selection for fenvalerate resistance was conducted in two colonies of the parasitoid, Cotesia plutellae, reared on two colonies of its host, Plutella xylostella, which differed in resistance to the insecticide. In the selection regime, the insecticide was applied to the host larvae that harboured the parasitoid larvae. Compared with the unselected parasitoid colony, the parasitoid colony selected with the susceptible hosts acquired 4.5 fold resistance after 14 selection cycles, while the colony selected with the more resistant hosts acquired 13.6 fold resistance after 13 selection cycles. These results demonstrate that parasitoid larvae could be exposed to insecticide selection via the hosts, selection with more resistant hosts could accelerate development of resistance in the parasitoid and resistance genes selected during larval development could be expressed at the adult stage. Comparison of detoxifying enzymes between the insect colonies revealed that fenvalerate resistance was positively related to increased monooxygenase activity, but was unrelated to carboxylesterases and general esterase in both the host and the parasitoid, indicating that the two insects shared a major metabolic mechanism for resistance to fenvalerate. This information can help improve selection procedures for the development of insecticide resistance in parasitoids.

Keywords

insecticide resistance, selection method, fenvalerate

Introduction

The use of naturally or artificially selected insecticide-resistant strains of natural enemies has been advocated to enhance the compatibility of biological and chemical controls (Croft 1990). Some natural enemy populations have developed high levels of resistance to insecticides in the field and can survive field application rates (Rathman et al. 1990, Baker & Weaver 1993). Many attempts have been made to achieve genetic improvement of natural enemies for insecticide resistance through laboratory selection (Johnson & Tabashnik 1994). The most successful cases are strains of predatory phytoseiid mites that can survive insecticide applications in the field (Hoy et al. 1983, Whitten & Hoy 1999). Using laboratory selection, azinphosmethyl resistance was increased 7.5 fold in the aphid parasitoid, Trioxys pallidus Haliday (Hoy & Cave 1989).

While examples exist to illustrate the potential of natural enemies to develop insecticide resistance in the laboratory and field, documented cases of insecticide resistance in field populations of natural enemies are relatively rare (Croft 1990). Likewise, laboratory selections for insecticide resistance in natural enemies have had limited success, particularly with insect parasitoids (Johnson & Tabashnik 1994). Many authors have recognised the need to improve selection methods for more successes (Rosenheim & Hoy 1988, Whitten & Hoy 1999).

To date, in all selection programs with parasitoids, the insects were exposed at the adult stage to insecticides, usually as residue inside glass vials or as a toxicant mixed in sugar solutions (Johnson & Tabashnik 1994, Li & Liu 2001). While this method provides a simpler procedure than selection of immature stages, it has an obvious disadvantage of exerting selection on, in most cases, only female phenotypes because mating usually precedes selection and male genomes are transmitted randomly with respect to insecticide resistance throughout a selection regime. This random transmission of male genomes prior to each selection cycle may delay the selection response. It is also possible that immature stages have different responses to selection from adults. A possible advantage for selection with immature stages is to explore the use of host resistance in the process, as parasitoid eggs and larvae develop inside the hosts.

In this study, we investigated the selection responses of the larval stage of the parasitoid, Cotesia plutellae Kurdjumov (Hymenoptera: Braconidae), to fenvalerate, using two host colonies that differed in susceptibility to the insecticide. Cotesia plutellae is a major parasitoid of the diamondback moth (DBM), Plutella xylostella (L.) (Lepidoptera: Plutellidae) (Talekar & Shelton 1993, Liu et al. 2000). Our purposes were to determine the feasibility of selection at the larval stage of the parasitoid and whether host resistance to an insecticide would favour selection of resistance in the parasitoid.

Materials and methods

Host and parasitoid colonies: rearing methods

A fenvalerate-susceptible stock culture of DBM, originally collected from Wuhan, Hubei, China, has been maintained on rapeseed seedlings since 1995. In March 1999, approximately 1600 pupae were collected from this culture and randomly divided into two groups to start a non-selected and a selected colony of DBM on potted cabbage respectively (Table 1). For convenience, these founding pupae are referred to as Generation 0 and their progenies as Generation 1, etc.

A laboratory stock culture of C. plutellae was initiated with approximately 200 cocoons collected from a cabbage farm in Hangzhou, China in 1995. Every year 100–200 newly field-collected cocoons were added to the culture. Fenvalerate has been one of the major insecticides used in the collection area since the mid 1980s. In February 1999, approximately 2100 cocoons were collected from this stock culture and divided randomly into three groups of approximately 100, 1000 and 1000 respectively. The first 100 were used to initiate a non-selected colony of the parasitoid using the non-selected colony of DBM as hosts. The next two groups were each provided with approximately 9000 II and III instar larvae from the non-selected host colony to initiate the selected with non-selected host (“Selected-NH”) and the selected with selected host (“Selected-SH”) colonies respectively (Table 1). As for the host colonies, these founding wasps are referred to as Generation 0 and their progenies as Generation 1, etc.

Table 1. Establishment and maintenance of selected and non-selected colonies of Plutella xylostella and Cotesia plutellae for artificial selection of resistance to fenvalerate

Colony	Details
A. Plutella xylostella
Non-selected	Initiated with 800 pupae from a fenvalerate-susceptible colony and maintained on cabbage plants without exposure to insecticides
Selected	Initiated with 800 pupae from the same fenvalerate-susceptible colony as for the Non-selected colony, maintained on cabbage plants and also used as the host for maintaining the Selected-SH colony of C. plutellae, selected with fenvalerate at the IV instar (together with the Selected-SH colony of C. plutellae) every generation for 13 generations
B. Cotesia plutellae
Non-selected	Initiated with 100 cocoons collected from a laboratory culture of C. plutellae and maintained using larvae from the Non-selected colony of P. xylostella as hosts without exposure to insecticides
Selected-NH	Initiated with 1000 wasps from the same laboratory colony as for the Non-selected colony, provided with II and III instar larvae of the Non-selected colony of P. xylostella as hosts, and selected at larval stage with fenvalerate every generation for 14 generations
Selected-SH	Initiated with 1000 wasps from the same laboratory colony as for the Non-selected colony, provided with II and III instar larvae of the Selected colony of P. xylostella as hosts, and selected at larval stage with fenvalerate every generation for 13 generations

All of the colonies were maintained on potted cabbage plants in stainless steel rearing cages in separate cubicles at 25±1°C, 14L:10D and 50–80%RH.

Insecticide

Fenvalerate, technical grade (85%, Hangzhou Pesticide Factory) was diluted with acetone for topical application to DBM larvae in bioassays. Fenvalerate, 20%EC (Jiangsu Red Sun Group Ltd) was diluted with distilled water to the desired concentrations when used in the resistance selection sprays and for the bioassays with the parasitoid adults.

Artificial selection

Selection was carried out when the parasitoids were in the larval stage inside host larvae. When host larvae developed to the II and early III instars, they were exposed to adult parasitoids for parasitism. Six days after exposure, all host larvae, either parasitised or not, reached the mid-late IV instar, and the parasitoid immatures inside were in the early to mid stages of larval development. The plants bearing the exposed hosts were then sprayed with a fenvalerate solution to run-off using a hand-sprayer. Host pupae and parasitoid cocoons were collected 4–5 days later to start the next generation cycle.

The selections for both the Selected-NH and Selected-SH colonies of the parasitoid were started with a spray of 200 mg ai/L fenvalerate, an approximate LC₅₀ for the non-selected colony of DBM. From the second generation onward, the parasitoid adults of the Selected-NH colony were always provided with host larvae collected from the Non-selected DBM colony for subsequent selection. For the selection of the Selected-SH colony, the moths that survived the previous selection were provided with host plants for oviposition to continue the Selected DBM colony (Table 1). Part or all of the host larvae, depending on the number available, were then exposed to the parasitoids that survived the previous selection. The fenvalerate concentrations were adjusted at each selection cycle so as to ensure some survivors to continue the colonies. Figure 1 shows the concentrations of fenvalerate used for the selections.

Figure 1. Concentrations of fenvalerate used in the selection regime for two colonies of Cotesia plutellae reared respectively in two colonies of Plutella xylostella that differed in resistance to the insecticide as the selection proceeded.

Bioassay of Plutella xylostella larvae

A series of five concentrations of fenvalerate diluted in acetone (10–90% mortality range) was made for the bioassay of each colony, in which five doses plus a control were each replicated three times with 10 larvae per replicate. One droplet of 0.5 μ1 fenvalerate solution was applied to the dorsal region of each larva using a Hamilton microsyringe. The larvae were reared in groups of 10 on cabbage leaf discs in 8 cm Petri dishes at 25°C, 14L:10D and mortality was assessed after 48 h.

Bioassay of Cotesia plutellae adults

For each C. plutellae colony, the susceptibility of adults to fenvalerate was assessed using a direct contact residual method with glass vials. For each bioassay, six concentrations (10–90% mortality range) of fenvalerate plus a control were prepared in distilled water. A fenvalerate solution of a given concentration was poured into each vial to its full capacity. After 10 s, the solution was poured off and the residue was air-dried at 20°C for 12 h, and three vials were treated with each concentration. Ten male and 10 female adults (12–24 h post emergence and fed with 10% honey solution) were introduced into each of the treated vials. The vials were capped with clean, untreated nylon gauze. No food was provided in the vials. The test vials were placed at 25°C, 14L:10D and mortality was assessed after 24 h.

Measurement of detoxifying enzyme activities

Activities of monooxygenase, carboxylesterases and general esterase in the various colonies of both the host and parasitoid were measured at the end of the selection regimes following the methods of van Asperen (1962) and Shang and Soderlund (1984). For the biochemical analysis of enzymes with non-parasitised DBM, IV instar larvae were used. For measuring enzyme activities in parasitised DBM and parasitoid larvae, parasitised DBM IV instar larvae harbouring mid-stage parasitoid larvae were dissected to remove parasitoid larvae, and the host larvae and the parasitoid larvae were then tested separately in the biochemical analysis. For enzyme analysis with parasitoid adults, male and female wasps (0–24 h after emergence) were used. Three samples, consisting of five individuals each, were analysed for each colony of the host and parasitoid.

Statistical analysis

Fenvalerate doses for bioassays with DBM larvae were translated to μg ai/mg body weight of test larvae for probit analysis. Dose or concentration-mortality data were analysed by probit analysis using POLO (LeOra Software 1997). Differences in susceptibility were considered significant when 95% confidence limits of LD₅₀s or LC₅₀s did not overlap.

Results

Selection of resistance

Selection was carried out over twelve months (March 1999 to March 2000). During this period, 14 generations of the Selected-NH colony and 13 generations of the Selected-SH colony were maintained. Selection was successfully applied to each generation of both colonies. The fenvalerate concentration used in selection for the Selected-NH colony was doubled from 200 mg ai/L to 400 mg ai/L from 2^nd to 5^th generation, but was reduced to 200 mg ai/L at the 6^th generation because of the low number of survivors. Thereafter, we were able to increase the selection pressure only at the 8^th and 9^th generation, and used 200 mg ai/L for most of the generations to the end of 14 selection cycles (Figure 1). By contrast, we were able to apply much higher selection pressure to several generations of the Selected-SH colony (Figure 1).

Bioassays for fenvalerate susceptibility for the host and parasitoid colonies were undertaken at the end of the selection. Differences between LD₅₀ values of the Non-selected and Selected DBM colonies were significant. Compared to the Non-selected colony, the resistance factor to fenvalerate of the Selected colony increased to 60 fold (Table 2). Differences between LC₅₀ values for adults of the three colonies of C. plutellae were also significant. Compared to the Non-selected colony, the resistance factors to fenvalerate of the Selected-NH and Selected-SH colonies increased 4.5 and 13.6 fold respectively (Table 3).

Table 2. Susceptibility of Plutella xylostella larvae to fenvalerate

Colony	N	Slope (SE)	LD50(95%CL) (μg/mg)a	χ2	df	RFb
Non-selected	150	0.790 (0.139)	0.103 (0.054–0.216)	1.24	3	1
Selected	150	0.998 (0.150)	6.244 (3.645–10.936)	1.98	3	60.3

^a μg ai/mg of body weight of test larvae. ^b RF, Resistance factor calculated by dividing the LD₅₀ of selected colony by LD₅₀ of the Non-selected colony.

Table 3. Susceptibility of Cotesia plutellae adults to fenvalerate

Colony	n	Slope (SE)	LC50 (95% CL) (mg ai/Litre)	χ2	df	RFb
Non-selected	240	0.865 (0.140)	4.1 (2.0–6.7)	0.63	2	1
Selected-NHa	300	0.833 (0.103)	18.4 (11.2–28.9)	1.49	3	4.5
Selected-SHa	300	0.673 (0.094)	55.7 (31.3–107.2)	2.78	3	13.6

^a Selected-NH = Selected with the Non-selected host colony; Selected-SH = Selected with the Selected host colony. ^b RF, Resistance factor calculated by dividing the LC₅₀ of the selected colonies by LC₅₀ of the Non-selected colony.

Detoxifying enzyme activities

In DBM larvae, the activities of monooxygenase in the Selected colony were higher than those in the Non-selected colony, whether the larvae were healthy or parasitised (Table 4). In the parasitoid, the activities of monooxygenase of larvae from the Selected-NH and Selected-SH colonies were 1.11 and 1.50 times that of the Non-selected colony (Table 4). Likewise, the activities of monooxygenase of parasitoid adults from the Selected-NH and Selected-SH colonies were 1.18 and 1.57 times that of the Non-selected colony (Table 4). In contrast, no increases in the activity of either carboxylesterases or general esterase were detected with increasing host or parasitoid resistance (Table 4).

Table 4. Detoxifying enzyme activities (μg product/mg protein/min, mean±SD) of healthy and parasitised larvae of Plutella xylostella, and larvae and adults of Cotesia plutellae

Insects	Monooxygenase	Carboxylesterases	General esterase
A. Plutella xylostella
Non-Selected, larvae, healthy	4.2 ± 0.46	19.4 ± 2.17	23.4 ± 2.13
Selected, larvae, healthy	4.8 ± 0.75	18.2 ± 0.79	28.7 ± 1.07
Non-Selected, larvae, parasitised	3.3 ± 0.90	14.1 ± 3.48	18.0 ± 4.03
Selected, larvae, parasitised	3.9 ± 0.90	14.8 ± 3.10	17.9 ± 1.45
B. Cotesia plutellae
Non-selected, larvae	5.4 ± 0.57	22.2 ± 0.33	25.7 ± 2.37
Selected-NH, larvae	6.0 ± 0.99	15.4 ± 2.87	18.8 ± 3.70
Selected-SH, larvae	8.1 ± 0.94	18.9 ± 6.65	26.5 ± 4.78
Non-selected, adults	4.9 ± 0.95	23.2 ± 7.36	29.7 ± 3.53
Selected-NH, adults	5.8 ± 0.86	19.1 ± 5.10	26.4 ± 3.76
Selected-SH, adults	7.7 ± 0.88	20.4 ± 3.74	27.3 ± 6.18

Discussion

In this study, no special effort was made to increase the frequency of the resistance allele in the initial parasitoid colony for selection by means of prior field selection or collections of material from different sites. Yet, the Selected-SH colony increased its resistance to fenvalerate 13.6 fold after only 13 selection cycles. This increase is impressive in view of the general slow and moderate increases in tolerance to insecticides recorded in past selection programs with parasitoids (Rosenheim & Hoy 1988, Croft 1990, Ke et al. 1991, Tabashnik & Johnson 1999). Equally impressive was the significantly higher level of resistance acquired by the Selected-SH colony than that acquired by the Selected-NH colony (Table 3). The results of this study thus provide new information for improving the methods for laboratory selection of insecticide resistance in parasitoids in several ways. First, the data demonstrated that parasitoids could be exposed to selection pressure at the larval stage, probably both within and outside the hosts prior to cocoon formation. It has been shown that an insecticide applied to a host larva can reach the parasitoid larva even when the primary resistance mechanism results in a lower rate of accumulation of toxicant in the host, e.g. enhanced degradation, and many metabolic breakdown products may also be toxic (Furlong & Wright 1993). However, direct evidence for the contact with, and/or consumption of, the insecticide by the parasitoid larvae in our case is yet to be obtained. It was likely that the parasitoid mature larvae were also exposed to the insecticide residue on the host larval surface and plant substratum during the period between egression from the hosts and pupation. Second, selection with resistant hosts could speed up the response by the parasitoids. The two host colonies were derived from the same source and their difference in susceptibility to the insecticide occurred in the selection process. Third, resistance genes selected during larval development could be expressed at the adult stage.

Compared to selection with parasitoid adults, application of an insecticide to the larvae enabled selection pressure to be exerted on the parasitoids before mating and thus would speed up concentrating resistance genes in the colony. However, direct comparison between selections with larvae and adults are needed to quantify the difference and to discern whether the parasitoid response to selection differs physiologically between the stages, in addition to the differences concerned with mating.

The use of host resistance in promoting response to selection by a parasitoid has a prerequisite that the host must have much higher tolerance to the insecticide than the parasitoid. This is feasible in most cases because it is usually practical to select highly resistant insect pests in the laboratory, which may then be used as hosts to select for a resistance level in the parasitoid sufficiently high to survive field application rates.

The mechanisms of host resistance to fenvalerate in promoting response to selection by C. plutellae are not entirely clear. One apparent factor seemed to be the higher selection pressure applied to parasitoids via more resistant hosts (Figure 1). Oxidative degradation has been shown to be a major metabolic mechanism in pyrethroid resistance in DBM (Hung et al. 1990). Comparison of detoxifying enzymes between the host or parasitoid colonies in this study demonstrated that fenvalerate resistance was positively related to increased monooxygenase activities, but was unrelated to activities of carboxylesterases and general esterase in both insects (Table 4), indicating that the two species shared a major metabolic mechanism for resistance to fenvalerate. It remains to be shown whether the acceleration of resistance selection in a parasitoid by the resistance in its host depends on the nature of resistance mechanisms.

Since a more resistant host harbouring a parasitoid larva would survive higher rates of the insecticide than a susceptible host and the parasitoid could develop to adulthood, host resistance can confer protection to endo-larval parasitoids such as C. plutellae (Furlong & Wright 1993, Iqbal & Wright 1996). Our results suggest that host resistance to an insecticide not only confers protection to the parasitoids, but also helps selection of resistance genes in the parasitoid. This information may offer new insights in understanding the development of resistance to insecticides by parasitoids in the field (Tabashnik & Johnson 1999).

Acknowledgements

This work was funded jointly by the China National Natural Science Foundation (Project No. 30070111) and the Australian Centre for International Agricultural Research (Project No. ACIAR CS2/1998/089). We thank a.m. Shelton and J.Z. Zhao, Department of Entomology, Cornell University, NYSAES, USA, and M.J. Furlong, Department of Zoology and Entomology, the University of Queensland, Australia for helpful comments on earlier versions of the article.

References