Sublethal Effects of Insecticides on the Parasitism of Acerophagus flavidulus (Hymenoptera: Encyrtidae) Parasitoid of the Obscure Mealybug, Pseudococcus viburni (Hemiptera: Pseudococcidae)

Abstract

Insecticides used to control agricultural pests can interfere with beneficial arthropods. This study determined the sublethal effects of two insect growth regulators—buprofezin and pyriproxyfen—and the neonicotinoid insecticide acetamiprid on adults of Acerophagus flavidulus (Brethés), a parasitoid of the obscure mealybug, Pseudococcus viburni (Signoret). A. flavidulus was exposed to insecticide residues at the minimum recommended rate of buprofezin and pyriproxyfen (1×) and 0.005× of acetamiprid on apple leaves under laboratory conditions. Each female parasitoid was in contact with the insecticide residues for 24 h and then allowed to parasitize three mealybug densities (two, four, and six nymphs) per parasitoid for 24 h. Parasitism, emergence rate, clutch size, development time, longevity, and secondary sex ratio were evaluated under each insecticide treatment and mealybug density. Application of the growth regulators buprofezin and pyriproxyfen at the labeled rate (1×) induced less sublethal effects than acetamiprid applied at a low rate (0.005×) on A. flavidulus. Pyriproxyfen and acetamiprid reduced parasitism, but they did not affect other aspects of development such as emergence rate, clutch size, development time, longevity, and secondary sex ratio. Our data suggest that buprofezin and pyriproxyfen are more compatible with A. flavidulus than acetamiprid, which could be integrated with parasitoid activity only when low residue levels in the field are attained.

1. Introduction

There is currently a high global demand for more environmentally sustainable agricultural products [1,2,3]. The increase in population growth has brought with it a disruption of agroecosystems, destabilizing them, which is generating a shift toward more sustainable practices that will ensure a more promising future for upcoming generations [4]. Thanks to methods based on ecological principles, such as biological pest control, the aim is to restore a certain balance between the populations of arthropods co-existing in agroecosystems. These natural enemies are used to prevent significant economic losses in crops affected by phytophagous pests [5]. These beneficial arthropods represent an alternative to the use of synthetic pesticides, presenting benefits such as less environmental damage and less risk to human health and, given the decrease in the availability of effective chemical products, they represent a very good source of pest control.

On the other hand, chemical control has been in the difficult process of finding new effective molecules for pest control, showing an increase in resistance [1,6], a rise in commercial regulations [7], and an increase in the concern of consumers [8,9,10]. Pesticides are considered to be one of the major drivers of the decline in biodiversity due to the exposure of non-target organisms in cultivated areas [10,11,12,13]. However, agricultural production around the world is challenged by several pests, which are managed mainly by chemical control [14]. As mentioned before, the increasing demand for agricultural goods requires pest management with an ecologically sustainable approach, which tries to include a rational use of synthetic pesticides combined with a balanced approach of different pest management strategies, such as the use of tolerant crops and biotechnology, among others [15,16], with biological control being one of the main ones. The biological control based on the release of natural enemies can contribute to more sustainable pest management, but they are frequently negatively affected by the use of broad-spectrum insecticides [17,18]. The objective then would be to find the sought-after balance between the use of biological control in association with selective synthetic pesticides, which represent the least possible damage to these natural enemies and the environment, using them only in cases where these are really necessary, ecologically and economically speaking. Therefore, the use of more selective and reduced-risk insecticides could improve the integration of chemical and biological control, and forms the basis for integrated pest management (IPM) strategies in crops.

Deciduous fruits are important crops in many temperate regions around the world. Pome, stone, and berry fruits as well as table grapes are produced for export markets in central Chile. All these crops are affected by several pests, among which the obscure mealybug, Pseudococcus viburni Signoret (Hemiptera: Pseudococcidae), is a main concern [19,20]. P. viburni is a widely distributed, polyphagous mealybug with quarantine-pest status in various countries [21,22,23]. Management of P. viburni in deciduous fruits is based mainly on chemical control. Broad-spectrum neurotoxic insecticides, such as acetamiprid, and other more selective insect growth regulators (IGRs), such as buprofezin and pyriproxyfen, are frequently used in deciduous fruit crops in Chile [24]. On the other hand, biological control of P. viburni is associated with predators (e.g., Coccinellidae and Chrysopidae) and parasitoid wasps (e.g., Encyrtidae) [25,26,27]. Among the parasitoid wasps, Acerophagus flavidulus Brethés (Hymenoptera: Encyrtidae) is the most important and efficient biocontrol agent for P. viburni [28,29]. In deciduous fruit crops, regular applications of broad-spectrum insecticides can interfere with the use of biological control [30,31,32]. Therefore, to develop effective IPM strategies, using both natural enemies and selective insecticides, the secondary effects of the latter on natural enemies must be considered [33,34,35]. In particular, it is very difficult for biological control to be compatible with the existing chemical control of obscure mealybug on deciduous fruit trees [24]. Thus, it is relevant to establish how long it would be necessary to wait for the inundative release of parasitoids after insecticide spraying. In our study, certain effects induced by sublethal doses of three insecticides used in P. viburni management—buprofezin, pyriproxyfen and acetamiprid—,will be evaluated to establish when A. flavidulus releases can be carried out.

This is important to study since sublethal effects of insecticides have been reported in several encyrtid parasitoids of mealybugs that affect their parasitism, emergence rate, development time, and longevity [6,36,37,38,39,40]. These effects affect their fitness and, therefore, the biological control of these mealybug pests.

The aim of this study is to evaluate the sublethal effects of a broad-spectrum insecticide (acetamiprid) and two more selective IGRs (buprofezin and pyriproxyfen) used in deciduous fruit crops on the P. viburni/A. flavidulus pest/parasitoid system. We exposed A. flavidulus parasitoid females to insecticide residues on apple leaves under laboratory conditions to evaluate the parasitism, emergence rate, clutch size, development time, longevity, and secondary sex ratio at increasing densities of their host, P. viburni.

2. Materials and Methods

2.1. Insect Material

Colonies of P. viburni were reared in butternut squash (Cucurbita moschata D.) inside plastic containers with orifices and blue cloth over the tops for ventilation and to avoid mealybug escapes. These containers were kept in a room chamber with controlled conditions at 23 ± 5 °C and 50 ± 10% RH and in complete darkness.

The parasitoid A. flavidulus was reared at Xilema SpA, a biological control company located in Quillota, Chile. Parasitoid mummies were sent to our laboratories and were maintained in a CONVIRON^® chamber at 25 ± 0.1 °C, 50 ± 5% RH, and a photoperiod of 16/8 L/D until the emergence of adults. A drop of 125 μL of the diet solution (organic honey, water, and agar) was placed on each petri dish to feed the newly emerged parasitoids for two to three days and to allow mating.

2.2. Application of Insecticides

The IGR buprofezin (Applaud™ 25 WP, Anasac, Santiago, Chile) was used at the minimum label rate recommended for obscure mealybug on apples (80 g·L⁻¹), and pyriproxyfen (Delico™ 100 EC, Anasac, Santiago, Chile) was applied at the label rate recommended for codling moth on apples (50 mL·L⁻¹). The neonicotinoid insecticide, acetamiprid (Hurricane 75 WP, Anasac, Santiago, Chile), applied at 0.005× of the label rate (0.6 mg·L⁻¹) to control P. viburni on apples, was tested too. These concentrations were defined as ”sublethal” concentrations in previous studies [24]. The residual route of exposure was selected because it tries to simulate the contact of the parasitoids after an insecticide spraying with the insecticide residues on leaf surfaces. A volume of 2 mL of each insecticide concentration was applied at 4.5 kPa with a Potter Precision Laboratory Spray Tower (BURKARD SCIENTIFIC, London, UK) on an apple-leaf disk, with the lower surface facing up on a pad of moist cotton inside a 30 mL plastic cup [41]. All the insecticides were diluted in distilled water, using distilled water as the control treatment. We did not use glass as a substrate for the application of insecticide residues, because it may overestimate the toxicological impact of the insecticides [42,43].

Once the treated surface had dried, one adult of A. flavidulus, which was ≤72 h old, was placed for 24 h in the bioassay arena, which had a lid with a small orifice (≈5 mm in diameter) on the top covered by MicroporeTM tape (3M^®, St. Paul, MN, USA). Every lid contained 125 μL of a diet for parasitoids. Plastic cups were maintained in a CONVIRON^® chamber with the same conditions previously described for parasitoid mummies. After this period, each parasitoid was placed in a 90 mm diameter plastic petri dish which contained the densities of two, four, and six third-instar nymphs of P. viburni. These mealybug densities were used since Karamaouna and Copland [25,44] reported that the oviposition capacity of A. flavidulus reaches a saturation point close to 2–5 mealybugs (nymphs) per female.

During 24 h, the parasitoid could oviposit onto the P. viburni offered without any diet. After this time, the parasitoid was killed to observe its terminalia under a stereoscopic microscope to identify the presence of an ovipositor. Sexing after the experiment was performed because it was difficult to manipulate and distinguish females before the experiment without producing damage. The parasitoid was fed from its emergence and for 24 h while it was in contact with the insecticides. The total time of the bioassay ranged from 4 to 5 days, during which the parasitoid survived under the above-mentioned conditions.

Petri dishes containing P. viburni were sealed again and maintained under the same controlled conditions in the chamber. The parasitism rate (number of mummies per number of mealybug nymphs offered) was measured on day 14, and mummies were individually separated into 0.5 mL Eppendorf tubes to await the emergence of parasitoids. In the lid of each Eppendorf tube, a 125 μL of diet was placed for feeding the newly emerged parasitoids. The sublethal effects were evaluated daily until the death of all individuals. The emergence rate (number of emerged mummies per number of mummies), clutch size (number of emerged parasitoids per mummy), and development time (days from oviposition to adult emergence) were measured.

Newly emerged parasitoids were separated into new Eppendorf tubes to determine their longevity and secondary sex ratio. The experimental unit was one female of A. flavidulus in a fixed density of P. viburni nymphs. The bioassays were replicated three times with the following number of experimental units per treatment: density of two mealybugs (9–14 experimental units), four mealybugs (12–18 experimental units), and six mealybugs (9–12 experimental units.

2.3. Statistical Analysis

Generalized linear models (GLM) with a binomial distribution were used to analyze data on the parasitism rate, emergency rate, and secondary sex ratio, and Poisson distribution was used to analyze the clutch size. Development time and longevity were analyzed using a two-way ANOVA because the assumptions of homoscedasticity and normality were fulfilled. The effects of insecticides, mealybug densities, and their interaction were evaluated. If the interaction was not significant, the additive model was used since it had lower Aikaike information criterion (AIC) values. If significant differences (p < 0.05) were found for the insecticide treatments or the mealybug densities, they were evaluated with the multiple-comparisons Tukey test, using the “multcomp” and “emmeans” R packages. All analyses were performed using the statistical software R version 4.1.0. [45].

3. Results

For the parasitism rate, a significant interaction was found between insecticide treatment and mealybug density (χ² = 12.66, df = 6, 11; p = 0.048). The parasitism rate at the host density of six mealybugs was lower for pyriproxyfen and acetamiprid in relation to the control (

Table 1. Mean parasitism rate, emergence rate, and clutch size of the parasitoid A. flavidulus, at three densities of P. viburni.

Insecticide	Mealybug Density		Parasitism Rate		Emergence Rate		Clutch Size
		Na	Mean ± SE	Nb	Mean ± SE	Nc	Mean ± SE
Control	two	54/66	0.82 ± 0.006	6/54	0.11 ± 0.002	26/6	4.33 ± 0.57
Buprofezin		41/58	0.71 ± 0.008	10/41	0.18 ± 0.004	43/10	4.30 ± 0.26
Pyriproxyfen		43/56	0.77 ± 0.008	8/43	0.19 ± 0.004	38/8	4.75 ± 0.42
Acetamiprid		52/84	0.62 ± 0.006	10/52	0.19 ± 0.004	26/10	2.60 ± 0.20
Control	four	111/164	0.68 ± 0.003	14/111	0.13 ± 0.001	45/14	3.21 ± 0.16
Buprofezin		134/216	0.62 ± 0.002	17/134	0.23 ± 0.002	74/17	4.35 ± 0.11
Pyriproxyfen		116/200	0.58 ± 0.002	17/116	0.15 ± 0.001	53/17	3.12 ± 0.10
Acetamiprid		89/152	0.59 ± 0.003	15/89	0.17 ± 0.002	34/15	2.27 ± 0.10
Control	six	116/162	0.72 ± 0.003 a	20/116	0.17 ± 0.001	68/20	3.40 ± 0.12
Buprofezin		148/216	0.69 ± 0.002 ab	9/148	0.15 ± 0.001	50/9	5.56 ± 0.35
Pyriproxyfen		95/180	0.53 ± 0.003 bc	2/95	0.02 ± 0.0002	9/2	4.50 ± 1.77
Acetamiprid		83/186	0.45 ± 0.003 c	16/83	0.19 ± 0.002	56/16	3.50 ± 0.14

N^a = number of parasitized hosts per total offered mealybugs (parasitism rate). N^b = number of emerged mummies per total parasitized hosts (emergence rate). N^c = number of emerged parasitoids (offspring) per total successful parasitized hosts. Different letters within each column indicate that the values are statistically significant (p ≤ 0.05).

). The main effects of insecticide treatment (χ² = 37.62, df = 3, 11; p ≤ 0.001) were significant, although the control and buprofezin were not significantly different, and both insect growth regulators did not show differences in parasitism rate. However, pyriproxyfen showed the same parasitism rate as acetamiprid. The mealybug density (χ² = 15.46, df = 2, 11; p ≤ 0.001) was significant too, and there was no difference between density four and six (p = 0.78) compared to density two.

For the emergence rate, the interaction between insecticides and mealybug density was also significant (χ² = 21.46, df = 6, 11; p = 0.002), but no significant differences between combinations of treatments were detected by the Tukey test (Table 1). The main effects of insecticide treatment (χ² = 6.81, df = 3, 11; p = 0.08) and mealybug density (χ² = 5.41, df = 3, 11; p = 0.07) were not significant.

For the clutch size, the insecticide treatment was significant (χ² = 19.75, df = 3, 11; p ≤ 0.001), with buprofezin showing a higher clutch size compared with acetamiprid (p ≤ 0.001) and the control (p = 0.02) (Table 1). Similarly, the clutch size was significantly affected by mealybug density (χ² = 6.26, df = 2, 11; p = 0.044). The density of six mealybugs increased the number of emerged parasitoids per mummy relative to the density of four (p = 0.04) but not for the density of two mealybugs (Table 1). No interaction between insecticide treatment and mealybug density was found (χ² = 5.83, df = 6, 11; p = 0.44).

For development time, only the insecticide treatment was significant (F = 3.25, df = 3, 11; p = 0.024), with buprofezin showing a shorter development time compared with the control (p = 0.04) (

Table 2. Development time, longevity, and secondary sex ratio of the parasitoid A. flavidulus, at three densities of P. viburni.

Insecticide	Mealybug Density		Development Time (days)	Longevity (days)		Sex Ratio (Males/Total)
		Na	Mean ± SE	Mean ± SE	Nb	Mean ± SE
Control	two	26	20.83 ± 0.27	3.72 ± 0.54	24	0.16 ± 0.04
Buprofezin		43	18.70 ± 0.09	4.13 ± 0.14	9	0.13 ± 0.03
Pyriproxyfen		38	19.72 ± 0.12	3.75 ± 0.22	25	0.34 ± 0.05
Acetamiprid		26	20.69 ± 0.17	3.71 ± 0.14	23	0.29 ± 0.04
Control	four	45	19.52 ± 0.09	4.64 ± 0.14	31	0.16 ± 0.01
Buprofezin		74	20.06 ± 0.10	4.55 ± 0.10	46	0.45 ± 0.03
Pyriproxyfen		53	20.00 ± 0.12	4.10 ± 0.10	19	0.26 ± 0.02
Acetamiprid		34	20.30 ± 0.14	6.06 ± 0.25	22	0.24 ± 0.03
Control	six	68	21.60 ± 0.13	4.15 ± 0.09	53	0.26 ± 0.02
Buprofezin		50	19.67 ± 0.24	4.90 ± 0.20	48	0.07 ± 0.01
Pyriproxyfen		9	20.00 ± 0.00	3.79 ± 0.86	8	0.00 ± 0.00
Acetamiprid		56	20.94 ± 0.13	5.16 ± 0.26	49	0.18 ± 0.02

N^a = offspring. N^b = offspring identified (male–female) in the presence or absence of an ovipositor.

). No effects of density of mealybugs (F = 2.45, df = 2, 11; p = 0.09) or interaction between the main factors was found (F = 1.51, df = 6, 11; p = 0.18).

Longevity was not affected by the insecticide treatment (F = 1.42, df = 3, 11; p = 0.24) or by mealybug density (F = 1.92; df = 2, 11; p = 0.15), and no interaction between the main factors was found (F = 0.47, df = 6, 11; p = 0.83) (Table 2).

There was no significant effect of the insecticide treatments (χ² = 0.38, df = 3, 11; p = 0.95) or interaction (χ² = 9.17, df = 6, 11; p = 0.16) on the secondary sex ratio. However, the mealybug density was significant (χ² = 8.04, df = 2, 11; p = 0.018), with densities of six and two mealybugs showing a lower number of males (p = 0.014) compared with the intermediate density of four mealybugs (Table 2).

4. Discussion

Three insecticides of different groups, widely used in deciduous fruits, were tested on the encyrtid parasitoid A. flavidulus. It has been reported that IGRs at the label rate are slightly harmful to A. flavidulus [24]. Nevertheless, in this study, the parasitism rate was not affected by buprofezin, but it was reduced with residues of pyriproxyfen at the highest mealybug density evaluated. Similar results were reported in Leptomastix dactylopii (Howard), a parasitoid of the citrus mealybug, Planococcus citri (Risso), where parasitism was reduced with pyriproxyfen at the label rate [36,37]. On the contrary, buprofezin applied at the label rate onto adults of L. dactylopii decreased parasitism and the emergence rate [37]. This could be explained by the different formulation (40 SC vs. 25 WP) and water volume (0.8 mL vs. 2 mL) used in our study, which resulted in a lower amount of active ingredient per parasitoid.

As mentioned by Suma et al. [39], the highest label rate of buprofezin and pyriproxyfen sprayed on adults of L. dactylopii did not affect the emergence rate or survival of females exposed to these insecticides for 72 h. In our study, buprofezin, besides being innocuous on the parasitism rate, emergence rate, and clutch size, led to a shorter development time on the P. viburni host than in the control. This situation was completely different when the same insecticide was applied to mummies of Planococcus ficcus (Signoret) parasitized by Anagyrus sp. near pseudococci (Girault) and Coccidoxenoides perminutus (Timberlake), both from encyrtidae family, where the development time was increased [38].

Wang et al. [46] reported that, compared with buprofezin, pyriproxyfen residues resulted in a significant reduction in parasitism on Encarsia formosa (Gahan), an endoparasitoid of Bemisia tabaci (Gennadius). A similar situation was observed in our study with a density of six mealybugs per parasitoid. In our bioassay, both insecticides were not statistically different, but pyriproxyfen as well as the neonicotinoid acetamiprid reduced significantly the parasitism rate. However, in E. formosa, pyriproxyfen, and buprofezin produced a significant decrease in parasitoid adult emergence rates, which did not occur in our study at any mealybug density. It is important to mention that parasitism was always lower as the host density increased, which may be because at low densities of P. viburni, the parasitoid did not have to make great efforts to parasitize it. The insecticides pyriproxyfen and acetamiprid decreased A. flavidulus parasitism on P. viburni at the highest density of mealybugs offered (six). If the number of hosts offered to each parasitoid is increased, parasitism may decrease further, which would be interesting to evaluate, since at a density of six mealybugs, A. flavidulus is close to the maximum oviposition capacity [25,44]. The sublethal effects of the three insecticides on the parasitism rate at the application rates evaluated were relatively low. In our experiments, the adult females had contact with the insecticide residues; therefore, developmental parameters in the offspring, such as the emergence rate, were not significantly affected. In all the studies previously mentioned, the sex ratio was not affected by either IGRs, as we have reported.

On the other hand, acetamiprid was harmful to A. flavidulus in terms of acute toxicity, even at 0.1× of the label rate [24]. However, at the reduced concentration tested of 0.005× of the label rate, it caused certain sublethal effects on this parasitoid, reducing its parasitism rate, especially at the highest mealybug density tested. This could be because acetamiprid is neurotoxic, and since it acts by ingestion and contact, the effect on the adult may have been more acute during the time it had contact with the insecticide residues, inducing a reduction in its ability to parasitize P. viburni. It has been shown that the uptake of neonicotinoids may impair the host-finding ability of females of Microplitis croceipes (Cresson), a parasitoid of the lepidopteran pest, Helicoverpa zea (Boddie), which occurs after they feed on extrafloral nectar from cotton plants [47]. Moreover, females of the encyrtid Avetianella longoi (Siscaro), a parasitoid of the cerambycid Phoracantha semipunctata (Fabricius), that were fed with accumulated imidacloprid in the nectar of treated eucalyptus trees had a significantly lower rate of parasitization [48]. For L. dactylopii, another formulation of acetamiprid (SG) at its label rate showed more detrimental effects on parasitism and the emergence rate but not on the sex ratio of the progeny [37].

Considering the above, IGRs at the minimum label rate would be more compatible with A. flavidulus than the neonicotinoid acetamiprid, especially buprofezin. This could be due to the rapid dissipation of the latter in apples, as demonstrated by González [49] with the same commercial product. Thus, buprofezin would be compatible with this parasitoid one day after treatment. Our results suggest that even at an acetamiprid concentration of 0.005×, parasitism of A. flavidulus is affected. Therefore, parasitoids should only be released once chemical residues are below sublethal levels. Considering the natural degradation of acetamiprid in apples and pears [49,50,51], a period of 20–25 days after chemical spraying of the insecticide is suggested before parasitoids can be “safely released” for inundative biological control. This estimation should be further evaluated because the insecticide dissipation data available in the previous references are from fruit and not leaves, which was the arena used in our bioassays.

This study provides evidence of the sublethal effects of selective insecticides, namely IGRs, and conventional insecticides such as neonicotinoids on the parasitoid A. flavidulus. To assess the risk of a pesticide on a natural enemy, it is essential to establish a link between the toxicity of the pesticide through laboratory tests and exposure under field conditions [35].

This will require the quantification of pesticide residues at sites where these beneficial arthropods may visit or inhabit along with the estimation of pesticide degradation in the field. In this way, risk assessment of the potential effects on key behavioral and physiological processes of these beneficial insects can be conducted, instead of only considering mortality as the most relevant endpoint, which is part of what was performed in this work.

Thus, it will be possible to identify specific side effects before the registration of pesticides for commercial use [35]. It is important to mention that this kind of laboratory results cannot be directly extrapolated to the field situations [52]. Studies combining field tests with laboratory assays found that, for a diverse set of mobile organisms, the population-level effects of insecticides in the field tend to be lower than would be expected based on laboratory results [53]. Nevertheless, further experimental studies linking laboratory data of sublethal insecticide effects with respective field data are critically needed to increase the credibility of risk assessments of insecticide use; these kinds of studies remain a major challenge [54]. Thus, the next step of this work would be to validate the compatibility of these insecticides against other parasitoid and predator species of the obscure mealybug under laboratory, semi-field, and field conditions to create an effective IPM program for P. viburni that allows for the use of certain insecticides that cause minimal impact on their beneficial arthropods. For future work, it is also important to target a broader scope and observe the effect of sublethal concentrations on insects’ life history and to expand this impact to a wider perspective, such as communities and the ecosystem.

From a population dynamics point of view, we found that the parasitism rate of A. flavidulus is sensitive to sublethal doses of pyriproxyfen and acetamiprid. Therefore, it could affect the biological control of P. viburni. These insecticides could directly affect the intrinsic growth rate (r_m) of the parasitoid through a lower oviposition rate and the subsequent lower mortality of the mealybug.

The study of the sublethal effects of insecticides on natural enemies is of great importance and needs to be considered when assessing the total effect of a toxicant [55]. All of the above-mentioned is relevant to understanding the potential interactions that may occur in order to achieve the desired pest control in different scenarios and to attain a sustainable agriculture system that does not compromise future generations and their needs [56].

5. Conclusions

This study determined the sublethal effects of three insecticides, the IGRs buprofezin and pyriproxyfen and the neonicotinoid acetamiprid used in deciduous fruit orchards on adults of A. flavidulus, a parasitoid of the obscure mealybug, P. viburni. The growth regulators buprofezin and pyriproxyfen, applied at the labeled rate, induced less sublethal effects than acetamiprid applied at a very low rate on A. flavidulus.

Acetamiprid at a very low concentration reduced parasitism, but it did not affect other aspects of development such as longevity, clutch size, and secondary sex ratio. This would indicate its compatibility with A. flavidulus only when very low residue levels are attained in the field.