Sublethal and transgenerational effects of dinotefuran on biological parameters and behavioural traits of the mirid bug Apolygus lucorum

The mirid bug, Apolygus lucorum, has become a major pest of many crops and fruit trees since the widespread adoption of Bt cotton in northern China. Neonicotinoid insecticides, such as dinotefuran, applied to control this pest may show sublethal effects, but evidence for such effects is lacking. Here, we investigated the sublethal and transgenerational effects of dinotefuran on biological parameters and feeding behavioural traits of A. lucorum using the age-stage, two-sex life table and electrical penetration graphs (EPGs), respectively. The LC10 and LC30 of dinotefuran against 3rd-instar nymphs of A. lucorum were 14.72 and 62.95 mg L−1, respectively. These two concentrations significantly extended the development duration from 3rd-instar nymph to adult in parent generation (F0). LC30 also increased the oviposition period and male adult longevity and reduced nymphal survival rate in the F0. For offspring generation (F1), the egg duration, preadult duration, and total preoviposition period were significantly lower at LC10 than in the control, and the egg duration, duration of 4th-instar nymphs, preadult duration, oviposition period, and fecundity were also decreased at LC30. However, the four demographic parameters of F1 generation, namely, net reproductive rate (R0), intrinsic rate of increase (r), finite rate of increase (λ), and mean generation time (T), were not affected by dinotefuran. The significant differences in the number of probes and duration of each of four feeding waveforms failed to be detected when A. lucorum nymphs treated by dinotefuran feed on Bt cotton plants without insecticide exposure. Overall, the dinotefuran concentrations tested here have sublethal, but no transgenerational impacts on A. lucorum.

Sublethal effects of dinotefuran on the F1 generation of Apolygus lucorum. The development duration and fecundity of F1 generation of A. lucorum are shown in Fig. 2. Both LC 10 and LC 30 significantly shortened the egg duration and preadult duration. Compared with the control, the duration of 4 th -instar nymph, oviposition period, and female fecundity were markedly decreased by LC 30 , and the total preoviposition period was also reduced by LC 10 . However, other biological parameters including the preadult survival rate (Control: 43 %; LC 10 : 50 %; LC 30 : 47 %) and the four demographic parameters, namely, R 0 , r, λ, and T, were not affected by these two concentrations ( Table 1).
The sublethal effects of dinotefuran on the age-specific survival rate (l x ), age-specific fecundity (m x ), age-specific maternity rate (l x m x ), and age-stage-specific fecundity (f xj ) of A. lucorum are shown in Fig. 3. The l x gradually decreased with ages and the maximum ages reached 76, 65, and 65 days in the control, LC 10 , and LC 30 , respectively. The time spans of female oviposition in the control were longer than that at the LC 10 and LC 30 (Control: 48 days, LC 10 : 37 days, LC 30 : 38 days). The m x has two peaks in each treatment across whole ages, and the f xj also has two peaks at LC 10 and LC 30 , but not in the control. The highest values of l x m x were 0.94, 0.92, and 0.78 for the control, LC 10 and LC 30 , respectively. Sublethal effects of dinotefuran on the feeding behaviour of Apolygus lucorum. Four main electrical waveforms were identified in A. lucorum fed on Bt cotton plants: stylet probing (P), stylet insertion into cells (I), cell rupturing and salivation (B), and feeding on the cell mixture (S). Neither LC 10 nor LC 30 significantly affected the number of probes or the duration of each waveform (P, I, B, and S) ( Table 2).

Discussion
In this study, we provided the evidence that both LC 10 and LC 30 of dinotefuran have no transgenerational effects on A. lucorum with respect to the demographic parameters: R 0 , r, λ, and T (Table 1). This finding was in accord with the previous reports on A. gossypii exposed to LC 20 of cycloxaprid 39 and B. tabaci treated by LC 25 of imidacloprid 40 . However, significant impacts on the demographic parameters were shown in A. lucorum treated with LD 15 of sulfoxaflor 41 ; in M. persicae exposed to imidacloprid 17,23 or thiamethoxam 20 ; and in A. gossypii 24 , Aphis glycines 42 and R. padi 43 exposed to imidacloprid. Zhen et al. 41 showed that the LD 15 of sulfoxaflor significantly reduced the r, λ, T, R 0 , and gross reproduction rate (GRR) of the F1 generation of A. lucorum compared with the control. Thus, the sublethal effects of insecticides on the demographic parameters were affected by multiple factors, e.g., insect species and kinds and amount of insecticides. Since the demographic parameters reflect the performance of insect pests at the population level, the present finding indicates that dinotefuran would not affect the performance of A. lucorum population.
The dinotefuran concentrations tested here also significantly increased the nymphal development duration, oviposition period, and male adult longevity in F0 generation of A. lucorum, while decreased the egg duration, preadult duration, total preoviposition period and oviposition period in F1 generation (Figs. 1 and 2). Interestingly, the oviposition period of A. lucorum exposed to LC 30 of dinotefuran in the F0 and F1 generation was inverse. Compensatory effects might exist in A. lucorum to synchronize the developmental rate of various populations. Similarly, Li et al. 43 observed a longer oviposition period in R. padi treated by imidacloprid, while a shorter oviposition period was documented in A. lucorum 18,41 and A. glycines 42 . Additionally, imidacloprid significantly extended the nymphal development duration of M. persicae 17,23 and R. padi 43 . In contrast, a reduction in nymphal development duration was reported in A. lucorum 41 , B. tabaci 15 , A. gossypii 24,34 , and A. glycines 42 . And the reduction in preadult duration may attribute to the lower egg duration at both LC 10 and LC 30 of dinotefuran in F1 generation of A. lucorum (Fig. 2). The development changes in these stages might be caused by two reasons. First, the antifeedant effects of these insecticides at low concentrations 22 negatively affected the nutrition absorption of exposed insects 22,24 . The other may be related to the disruption of hormone balance 42 .
We also showed that the nymphal survival rate in F0 generation and fecundity in F1 generation of A. lucorum were decreased at LC 30 . Many insect species also have lower survival rates in their immature stages, including A. lucorum 41 , R. padi 43 , B. tabaci 15,44 , Euschistus heros 16 , and A. gossypii 45 . Additionally, the reduction in fecundity were also observed in A. lucorum with LD 40 of cycloxaprid 18 , A. gossypii exposed to LC 10 and LC 40 of cycloxaprid 34 , A. glycine with 0.20 mg L −1 of imidacloprid 42 , and N. lugens treated with imidacloprid and dinotefuran 46 . This phenomenon could attribute to the reduction in vitellogenin (Vg) and the expression level of Vg mRNA significantly decreased by 43.8% in F1 generation of A. lucorum whose parents were treated with LD 15 of sulfoxaflor 41 . EPGs analysis demonstrated that the feeding behaviour of A. lucorum did not differ between dinotefuran treatments and the control, indicating that these two concentrations will not increase crop injury when this pest moves to other plants. In contrast, many studies have reported negative effects of insecticides on the feeding behaviour of A. gossypii 24,34 , M. persicae 23,35 , S. avenae 25 , and R. padi 27,36 . Cira et al. 26 found that H. halys adults that www.nature.com/scientificreports www.nature.com/scientificreports/ survived from sulfoxaflor exposure produced significantly fewer feeding sites than those in the control. Indeed, A. lucorum performs a macerating or lacerating behaviour, which is different from other sap-feeding pests such as aphids and leafhoppers [28][29][30][31] , and has a shorter ingestion duration like Adelphocoris suturalis in plants 47 . The  www.nature.com/scientificreports www.nature.com/scientificreports/ different feeding strategy may led to their different responses towards insecticides. Nevertheless, further experiments are needed to clarify the relationships between the amount of insecticide used and the feeding response of A. lucorum.
In summary, the dinotefuran concentrations tested here showed sublethal effects, but no transgenerational effects on A. lucorum. It implies that dinotefuran would not increase the population size of A. lucorum.

Materials and Methods
Ethics statement. Permission was not required for insect collection, because none of the species used in the study were endangered or protected.
Insects. Overwintering eggs of A. lucorum used in this study were originally collected from a winter jujube orchard in Binzhou, Shandong Province, China, in 2016. After eggs hatched, they were reared on green bean (Phaseolus vulgaris) without exposure to any insecticide in transparent glass jars (10 cm in diameter, 15 cm in height). These jars were maintained in a climate-controlled chamber with a temperature of 25 ± 1 °C, relative humidity (RH) of 65 ± 5%, and photoperiod of L16: D8.
Insecticides. Dinotefuran (95.57% purity) was purchased from Suzhou Aotelai Chemical Group (Suzhou, Jiangsu Province, China) and used in all the following experiments.   Table 2. Sublethal effects of dinotefuran on probing number and probe duration of Apolygus lucorum fed on Bt cotton plants for 6 h. P waveform represents stylet probing; I waveform represents stylet insertion into cells; B waveform represents cell rupturing and salivation; and S waveform represents feeding on the cell mixture. (2020) 10:226 | https://doi.org/10.1038/s41598-019-57098-z www.nature.com/scientificreports www.nature.com/scientificreports/ Acute toxicity of dinotefuran to Apolygus lucorum. The acute toxicity of dinotefuran to 3 rd -instar nymphs of A. lucorum was assessed using the leaf-dipping method. The stock solution of dinotefuran prepared in acetone was diluted into a series of concentrations: 68.75, 137.5, 275, 550, and 1100 mg L −1 , using distilled water containing 1‰ (v/v) Tween-80. Distilled water containing 1‰ Tween-80 was used as the control. Fresh green beans with the same size were cut into 2-cm-long sections, dipped into each insecticide solution and the control for 20 min, and air-dried at room temperature for 2 h. These beans were then placed into the transparent plastic containers (6 cm in diameter, 7 cm in height), each of which included three 2-cm-long sections of green bean. After 5 h of starvation, 3 rd -instar nymphs of A. lucorum were transferred into these containers. Each treatment was repeated three times with 15 nymphs per replicate. All the containers were then maintained in a climate-controlled chamber with a temperature of 25 ± 1 °C, RH of 65 ± 5%, and photoperiod of L16: D8. After 48 h, the nymphal mortality was recorded, and the nymphs that did not move when touched with a thin brush were regarded as dead.
Sublethal effects of dinotefuran on the F0 generation of Apolygus lucorum. The LC 10 (14.72 mg L −1 ) and LC 30 (62.95 mg L −1 ) were used to evaluate the sublethal effects of dinotefuran on A. lucorum. These two concentrations were prepared as the method described in the "Acute toxicity of dinotefuran to Apolygus lucorum" section above. Distilled water containing 1‰ Tween-80 was used as the control. There were 154, 135, and 225 of 3 rd -instar nymphs used for the control, LC 10 , and LC 30 , respectively. After 48 h, the survivors of each treatment were individually transferred to the smaller transparent plastic container (1.5 cm in diameter, 2 cm in height) with one 1.5-cm-long green bean section free from insecticide. The development and survival of nymphs were recorded daily. After adults emerged, they were paired (1 male and 1 female) in new transparent plastic containers (1.5 cm in diameter, 2 cm in height) with one 1.5-cm-long section of green bean (as food and oviposition substrate). The old green beans were replaced by new ones daily. The eggs in the old green beans were checked and counted under a stereomicroscope until adult death. If the male adult died during the experiment, it was removed and replaced by a new one from the same treatment. All the experiments were conducted in a climate-controlled chamber with a temperature of 25 ± 1 °C, RH of 65 ± 5%, and photoperiod of L16: D8.
Sublethal effects of dinotefuran on the F1 generation of Apolygus lucorum. When the eggs laid by female adults of the F0 generation peaked, there were 117, 131, and 134 eggs randomly selected and assigned to the control, LC 10 , and LC 30 , respectively, to initiate the life table study of the F1 generation. The hatched eggs were recorded daily and the newly born nymphs were individually transferred into a new transparent plastic container (1.5 cm in diameter, 2 cm in height) with one 1.5-cm-long green bean section without exposure to any insecticide. The nymphal stage and survival were checked and recorded daily. Within 24 h of adult emergence, adults were paired (1 male and 1 female) in a transparent plastic container with one 1.5-cm-long section of green bean (as food and oviposition substrate). The old green beans were replaced by new ones daily. The eggs in the old green beans were checked and counted under a stereomicroscope until adult death. If the male adult died during the experiment, it was removed and replaced by a new one from the same treatment. All the experiments were conducted in a climate-controlled chamber with a temperature of 25 ± 1 °C, RH of 65 ± 5%, and photoperiod of L16: D8.
Sublethal effects of dinotefuran on the feeding behaviour of Apolygus lucorum. The feeding behaviour of A. lucorum exposed to dinotefuran on Bt cotton plants at the seedling (variety: Lumianyan 36, developed by Cotton Research Center, Shandong Academy of Agricultural Sciences, Jinan, China) were recorded using a Giga-4 direct-current electrical penetration graph system (DC-EPG, manufactured by WF Tjallingii, Wageningen University, Wageningen, Netherlands). Briefly, the 3 rd -instar nymphs reared on green beans were exposed to LC 10 and LC 30 of dinotefuran and the control prepared as described in the "Acute toxicity of dinotefuran to Apolygus lucorum" section above for 48 h. After 5 h starvation, these nymphs were immobilized on an ice plate and then secured on a vacuum device for the attachment of wires. A gold wire (2 cm in length, 20 μm in diameter) was attached to the pronotum of individual nymph using silver glue under a stereomicroscope. The gold wire allowed the nymphs to move relatively unaffectedly with a radius equal to the length of the wire tether. Then, each nymph was connected to the amplifier before being placed on a cotton leaf. Another copper electrode was inserted into the soil in the pot with one Bt cotton plant. Finally, the entire experimental arena was covered by a Faraday cage to shield external noise and other interference. Recordings were made simultaneously on four individual Bt cotton plants over 6 h using ANA 34 software (Wageningen, The Netherlands). Nymphs and Bt cotton plants were used only once and then discarded. For each treatment, at least 20 nymphs were successfully tested. Electrical signals were identified and characterized based on the descriptions by Song et al. 33 and Zhao et al. 32 .
Data analysis. The LC 10 , LC 30 , and LC 50 values were determined based on the probit analysis. The development and fecundity of F0 generation and feeding behaviour were analysed using one-way analysis of variance (ANOVA) followed by Tukey's multiple-range test. The nymphal survival rate of F0 generation were compared by Chi-squared test (χ 2 ). All these analyses were conducted in SPSS 19.0 software (IBM Inc., New York, USA).
The life table data for all A. lucorum individuals in the F1 generation were analysed using TWOSEX-MSChart computer program 48 according to the age-stage, two-sex life table theory 49,50 . The demographic parameters, namely, the net reproductive rate (R 0 ), intrinsic rate of increase (r), finite rate of increase (λ), and mean generation time (T), of the F1 generation were calculated based on the following Eqs. (1-4) via the computer program. The age-specific survival rate (l x ), the age-specific fecundity (m x ), age-specific maternity rate (l x m x ), and the age-stage-specific fecundity (f xj , where x is the age and j is the stage) were also obtained from the computer program. The means and standard errors of all the life history traits and demographic parameters were estimated using bootstrap technique with 100,000 resamplings 51,52 , and the differences among treatments were compared using the paired bootstrap test based on the confidence intervals 53 . All the figures were created in SigmaPlot 14.0 software (Systat Software Inc., San Jose, CA). For all the test, α = 0.05.