Cabbage stem flea beetle’s (Psylliodes chrysocephala L.) susceptibility to pyrethroids and tolerance to thiacloprid in the Czech Republic

The cabbage stem flea beetle (CSFB), Psylliodes chrysocephala (Coleoptera: Chrysomelidae), has recently become a major pest species in winter oilseed rape in the Czech Republic. The susceptibility of CSFB populations from two localities to six pyrethroids, two neonicotinoids, one organophosphate and one oxadiazine was evaluated in 2015–2018 in glass vial experiments. The susceptibility of CSFB to thiacloprid and thiamethoxam was evaluated in feeding experiment in 2017 and 2018. High susceptibility of CSFB populations to lambda-cyhalothrin, cypermethrin, esfenvalerate, tau-fluvalinate, etofenprox, deltamethrin, chlorpyrifos, indoxacarb and acetamiprid was observed in the glass vial experiments. The LC50 and LC90 data obtained for pyrethroids in these experiments in 2015 represent baseline for CSFB resistance monitoring to pyrethroids in the Czech Republic. High tolerance of CSFB to thiacloprid of CSFB was demonstrated in glass vial and the feeding experiment, too.


Introduction
The cabbage stem flea beetle (CSFB), Psylliodes chrysocephala Linnaeus 1758 (Coleoptera: Chrysomelidae) is one of the major pest species of winter oilseed rape in Western Europe [1,2,3]. CSFB has one generation per year and undergoes summer diapause. The period of adult summer diapause is not influenced through constant or changing environmental temperatures or short or long daylight conditions [4]. In autumn, adults migrate into new fields of oilseed rape. Females lay eggs after mating on or close to the plants. After hatching, the larvae feed on leaf bases and terminal leaflets. All the developmental stages can overwinter and females can resume laying eggs in the spring.
Since 2009, the first evidence of CSFB resistance to pyrethroids has been demonstrated as a consequence of the long-term selective pressure of pyrethroids applied for the control of CSFB in oilseed rape [2]. The resistance factor according to the LC50 values for lambda-cyhalothrin PLOS  reached 80 when 28 local populations of CSFB from Germany were examined [1]. Failure to control CSFB has become a major issue in Germany and the United Kingdom [3]. Until now, no data were available about the susceptibility of CSFB to insecticides in the Czech Republic. Since 2013, three active neonicotinoid ingredients, thiamethoxam, clothianidin and imidacloprid, have been restricted in Europe for seed treatments [5]. Until the start of the restriction, CSFB control in northern Germany and further regions of Western Europe has relied on seed dressings with neonicotinoids or spraying with pyrethroids in autumn [2,6]. Similarly, CSFB in the Czech Republic was effectively controlled until 2013 by seed dressings of winter oilseed rape with neonicotinoids. Using pyrethroids in oilseed rape in autumn was sporadic. In 2014, the first economically important injuries to oilseed rape caused by CSFB were recorded by farmers in the central part of the Czech Republic. Since 2015, CSFB has been controlled in most of the regions of the Czech Republic by spraying with pyrethroids.
The present study brings the results of i) evaluation of the susceptibility of local populations of CSFB from the Czech Republic to six pyrethroids, two neonicotinoids, one organophoshate and one oxadiazine in glass vial experiment, ii) evaluation of susceptibility of CSFB to two neonicotinoids in feeding experiments.

Sample collection
The field populations of CSFB originated from winter oilseed rape fields located in Potěhy (49˚52 0 9.695@N, 15˚25 0 12.344@E) and Prague (50˚5 0 15.437@N, 14˚17 0 58.980@E). CSFB adults were collected by sweeping during June in 2015-2018. According to the CSFB population density, 500-1,500 adults per population were collected in particular years and populations. Beetles with some oilseed rape plant material in the net insulators were transported to the laboratory. Prior to initiating the bioassays, samples of the beetles were stored at +5˚C. The transported and stored beetles were fed winter oilseed rape leaves. The beetles were tested for susceptibility within 48 h after collection. Only live, mobile insects were used for the bioassays. A total of 4,876 CSFB beetles were tested in bioassays in 2015-2018.

Bioassays
Insecticides. Six analytical standards (lambda-cyhalothrin, cypermethrin, esfenvalerate, tau-fluvalinate, etofenprox and deltamethrin), one organophosphate (chlorpyrifos), one oxadiazine (indoxacarb) and two neonicotinoids (thiacloprid and acetamiprid) were obtained from Sigma Aldrich Co. LLC., Prague, Czech Republic. Four to seven doses of insecticides diluted with acetone were used to generate dose-response curves. The rates recommended in the Czech Republic for the field application were calculated assuming that the highest recommended rate of a commercial product in 400 L of water is used per hectare (Table 1). To generate dose-response curves, the rates of the tested insecticides corresponding to 0.1%-1,000% of the field rates were tested (Table 1). Thiacloprid was tested in 2015-2017 as the analytical standard and in 2017-2018 as the formulated product Biscaya 240 OD. Stock solutions of Biscaya 240 OD were prepared by dissolving 140.4 mg of Biscaya 240 OD formulation (containing 32.4 mg of thiacloprid) in 2 mL of distilled water, and subsequently adjusted to 100 mL with acetone. All further dilutions were made in acetone.
Glass vial experiment. IRAC adult vial tests [7] no. 31 (pyrethroids), no. 21 (neonicotinoids), no. 25 (organophosphates) and no. 27 (oxadiazines) were used to conduct the bioassay. The glass vials (P-lab, CZ, internal surface 32.4 cm 2 ) were filled with 520 μL of insecticide and rotated on a hotdog roller at room temperature until the acetone was completely evaporated. Subsequently, the beetles were transferred into each vial. Three replicates were used, and 10 beetles were used for each insecticide dose in each replication. The control vials were treated with pure acetone. The vials were incubated under controlled conditions at 20˚C and 60% relative humidity with a 16:8 h light:dark photoperiod. The number of severely affected beetles (i.e., dead and moribund) was counted after 48 h.
Feeding experiment. Oilseed rape of the Exssence variety was sown at the rate of 400,000 seeds per ha with a plot seeder Wintersteiger (Wintersteiger Sägen GmbH, Arnstadt, Germany) on small plots (0.2 ha) in the field in Prague locality with three variants: (1) without seed dressing, (2) seed dressing with Sonido (thiacloprid), and (3) seed dressing with Cruiser (thiamethoxam). The plots were left without treatment with herbicides or insecticides. The conditions of the experiment permitted the natural infestation by CSFB adults. The incidences of adults in the plots were monitored by Moericke yellow traps. The traps were inspected in 2-3 day intervals. In 2017, the plants from the experiment were collected in two periods, 24 days and 35 days after sowing of the oilseed rape seeds. 5.5 and 6.2 adults were caught per trap from the plant emergence to the first and second term of plant collection, respectively. Growth stage of oilseed rape plats according to BBCH-scale corresponding to dates of plant sampling are presented in Table 2. In 2018, the plants were collected 29 days after sowing of the oilseed

Analysis of residues of neonicotinoids in oilseed rape plants
The residues of thiacloprid and thiamethoxam as active substances of the seed dressings Sonido and Cruiser, respectively, were analysed in samples of oilseed rape plants by the QuE-ChERS method followed by LC-MS/MS. From each variant and repetition, a composite sample of 2.50 g of plants was used for the analyses. The analyses were carried out in the accredited laboratory at the University of Chemistry and Technology in Prague as a service.

Data analysis
The bioassay data were analysed using a probit analysis via a dose-effect analysis in XLSTAT 2015 (Addinsoft USA, New York, NY, USA). The doses of insecticides were ln-transformed. The lethal concentrations (LC50s) were fitted with 95% confidence limits. The LC50 values in particular bioassays were considered significantly different when the respective 95% confidence limits did not overlap. A Pearson correlation was used to analyse the correlation between years and LC50s of insecticides lambda-cyhalothrin, tau-fluvalinate and thiacloprid.
Pearson correlation was used also to analyse the correlation between years and slopes of the probit models calculated for insecticides lambda-cyhalothrin, tau-fluvalinate and thiacloprid. The mortality data of the CSFB after application of thiacloprid in particular years and localities in the glass vial experiments and the injury of plants, weight of the oilseed rape plants and content of active substances of pesticides in the seed-dressing experiments were compared using a one-way analysis of variance (ANOVA) after testing their normality by the Shapiro-Wilk test using the XLSTAT 2015 statistical software package (Addinsoft Inc., New York, USA). The means were separated using the Tukey HSD analysis.

Susceptibility to pyrethroids in the glass vial experiment
The susceptibilities of CSFB populations from Prague and Potěhy localities in 2015-2018 to 6 active substances of pyrethroids expressed as LC50s are given in Table 3. The probit data (N, LC50, slope, chi-square and R 2 ) for particularly years are presented in S1-S4 Tables. In relation to data from Mecklenburg-Western Pomerania in Germany [1], the LC50 values demonstrated that both tested local CSFB populations were highly susceptible to the pyrethroids cypermethrin, deltamethrin, esfenvalerate, etofenprox, lambda-cyhalothrin and tau-fluvalinate. The LC50s for the pyrethroids evaluated after 48 h ranged from 0.0002 μg/cm 2 (deltamethrin, Prague 2016) to 0.13 μg/cm 2 (tau-fluvalinate, Prague 2018). The LC50 values for lambda-cyhalothrin didn´t differ between years 2015-2018 in CSFB populations from both the localities ( Table 3). The LC50 values for tau-fluvalinate for CSFB from Potěhy in 2018 was significantly higher than the values in 2015 (Table 3). Similarly, the LC50 values for tau-fluvalinate in Prague in 2017 and 2018 were higher that the values in 2015 (Table 3). However, the slope of the bioassay with tau-fluvalinate on CSFB samples from Prague locality in 2015 was steep and calculation of confidence limits was not possible. Hence, comparisons of the CSFB susceptibility between 2015 and following years was not possible. The LC50 values for tau-fluvalinate in 2017 and 2018 in Prague locality were not significantly different ( Table 3).
The mortalities of CSFB from Prague and Potěhy localities after application of deltamethrin, cypermethrin, esfenvalerate, etofenprox, lambda-cyhalothrin and tau-fluvalinate in concentrations corresponding to 100% of the recommended field concentration were 100% in 2015. In 2017 and 2018, the mortality of CSFB from Prague locality after application of tau-fluvalinate in concentration corresponding to 100% of the recommended field concentration decreased to 86.7% and 79.2%, respectively (Table 4). However, the difference in CSFB mortalities from Prague locality between years was not statistically significant. Similarly, the mortality of CSFB from Potěhy decreased from 2015 to 2018 to 86.9% and the difference in mortalities was statistically significant (Table 4). No decrease of CSFB mortality in 2017 or 2018 was observed after application of lambda-cyhalothrin in both the localities ( Table 4).
The increased LC50 values for tau-fluvalinate in 2017 and 2018 together with the decreased mortality after application of tau-fluvalinate in 2017 and 2018 indicates a shift in CSFB susceptibility to tau-fluvalinate after three years of CSFB control by spraying tau-fluvalinate in the spring period onto oilseed rape.

Susceptibility to organophosphates in the glass vial experiment
Sufficient susceptibility of CSFB from two localities to chlorpyrifos was demonstrated in 2015-2018 (Table 3,   Cabbage stem flea beetle's (Psylliodes chrysocephala L.) susceptibility and tolerance The CSFB mortality after application of chlorpyrifos in concentration corresponding to 100% of the recommended field concentration was 100% in Prague locality ( Table 4). The same CSFB mortality was obtained in Potěhy locality in 2015 (data not presented). Data obtained for chlorpyrifos in 2015-2018 demonstrate stable high CSFB susceptibility from Prague to organophosphates.

Susceptibility to oxadiazine in the glass vial experiment
The LC50 values for CSFB from two localities in 2015-2016 showed a high susceptibility to indoxacarb (Table 3, S1 and S2 Tables). The LC50s for indoxacarb ranged from 0.02 μg/cm 2 (Prague 2016) to 0.03 μg/cm 2 (Prague 2015, Potěhy 2015) and did not differ significantly between populations. The CSFB mortality after application of indoxacarb in concentrations corresponding to 100% of the field concentration was 100% in 2015 in both the Prague (Table 4) and Potěhy (data not presented) localities. In 2016, the mortality of CSFB from Prague decreased to 97.5% (Table 4). However, the difference in mortalities between 2015 and 2016 was not significantly different.

Susceptibility to neonicotinoids in the glass vial experiment
The LC50 values indicate that the CSFB populations from two localities were susceptible to acetamiprid. The LC50s ranged from 0.002 μg/cm 2 (Prague 2016) to 0.03 μg/cm 2 (Prague 2015) ( Table 3, S1 and S2 Tables). The CSFB mortalities after application of acetamiprid in concentrations corresponding to 100% of the recommended field concentration was 100% in both the Prague (Table 4) and Potěhy (data not presented) localities.
In contrast to acetamiprid, the susceptibility of CSFB populations from both the localities to thiacloprid tested as analytical standard was insufficient. The LC50 values for thiacloprid ranged from 0.63 μg/cm 2 (Prague 2015) to 139 μg/cm 2 (Prague 2017) ( Table 3, S1-S3 Tables). In both localities, comparison of the LC50 values between the years showed a decreasing susceptibility of CSFB to thiacloprid from 2015 to 2017 (Table 3). The LC50 values obtained for thiacloprid as a formulated product Biscaya 240 OD were much lower that the values for analytical standard of thiacloprid ( Table 3).
The CSFB mortality after application of analytical standard of thiacloprid in concentrations corresponding to 100% of the recommended field concentration decreased from 50.0% in Prague locality in 2015 to 0% in both the localities in 2017 (Table 4). However, the difference was not statistically significant. The mortality data obtained for formulated product Biscaya 240 OD showed decreasing mortalities of CSFB from 2017 to 2018 in both the localities.

Susceptibility of CSFB to neonicotinoids in the feeding experiment
The content of pesticide residues differed between the treatment variants in both plant collection periods in 2017. In the first period (24 days from seeding), the content of thiamethoxam in plants was 17 times higher than the content of thiacloprid. In the second treatment period (35 days from seeding), the content of thiamethoxam in plants was 5 times higher than the content of thiacloprid ( Table 2). In 2018, 29 days from seeding, the content of thiamethoxam in plants was 2 times higher than the content of thiacloprid. The average weight of the plants was significantly higher in the thiamethoxam variant than in the thiacloprid and control variants in the first plant collection period in 2017. In the second plant collection period in 2017 and in 2018, the average weight of plants did not differ significantly between the variants thiacloprid and thiamethoxam. The average injury to plants in the thiacloprid variant did not differ significantly from the control variant without treatment and was significantly higher than in the thiamethoxam variant in both years and all periods of plant collection ( Table 2).

Susceptibility to pyrethroids in the glass vial experiment
The present study showed good efficacy of the pyrethroids lambda-cyhalothrin, deltamethrin, High susceptibility to pyrethroids with very low resistance ratios were also observed in Danish CSFB samples [3]. In contrast to this, evidence for kdr-resistance in CSFB was published from Germany [2], where the pyrosequencing diagnostic assay showed a kdr-allele frequency of 90-100% in populations expressing high levels of resistance in adult vial tests. Cross-resistance to lambda-cyhalothrin, tau-fluvalinate, etofenprox and bifenthrin was also shown [2]. In contrast to the German samples, where the survival of CSFB at high concentrations of lambdacyhalothrin correlated well with the presence of the kdr-allele, the presence of kdr-genotypes in samples from the United Kingdom did not always correlate well with resistant phenotypes. This indicates a possible incidence of an as yet undisclosed, metabolic-based mechanism [3]. Regarding an alarming trend in the spread of CSFB resistance to pyrethroids, including the kdr-mutation and metabolic-based mechanisms, it is not recommended to alternate the insecticides on the basis of pyrethroids in anti-resistance strategies. The introduction of pesticides with alternate modes of action is necessary.

Susceptibility to organophosphates and oxadiazine in the glass vial experiment
Evaluation of the susceptibility of CSFB to insecticides showed a high susceptibility of two local CSFB populations from the Czech Republic to chlorpyrifos and indoxacarb. Up to now, no resistance to chlorpyrifos has been recorded in CSFB. LC50 value 0.01 μg/cm 2 for chlorpyrifos was established in this paper for the most sensitive sample of CSFB population. Good efficacy of organophosphates was also published against Brassicogethes aeneus [9] and Ceutorhynchus obstrictus [10]. Chlorpyrifos has recently been recommended for the control of CSFB in the Czech Republic because of the need for alternation with pyrethroid-based insecticides in anti-resistance strategies.
Similarly, no resistance of CSFB or B. aeneus to indoxacarb was recorded up to now. LC50 value 0.02 μg/cm 2 for indoxacarb was established in this paper for the most sensitive sample of CSFB population. Till present, indoxacarb has not been used for control of CSFB in the Czech Republic and is a promising pesticide for anti-resistant strategies to slow down selection of CSFB resistance to pyrethroids. No cross-resistance between indoxacarb and pyrethroids and organophosphates was shown in insecticide-resistant Spodoptera frugiperda and Plutella xylostella [11]. However, a high resistance level to indoxacarb has been published for C. obstrictus from Poland [12].

Susceptibility to neonicotinoids in the glass vial experiment
This study showed different susceptibilities of both CSFB populations to the neonicotinoids acetamiprid and thiacloprid. Both tested populations of CSFB were highly susceptible to acetamiprid and simultaneously exhibited tolerance to thiacloprid. In the Czech Republic, thiacloprid is not registered for use against CSFB in oildseed rape crops. However, thiacloprid was frequently used in the Czech Republic against Ceutorhynchus obstrictus and Dassineura brassicae in flowering period of oilseed rape. These applications influence the CSFB larvae escaping the plants after termination of development. Hence, selection pressure of thiacloprid on CSFB larvae in not eliminated.
The slope values for probit models differed according to groups of insecticides. The lowest slopes with average value < 1.00 were calculated for thiacloprid, the highest slopes were calculated for chlorpyrifos with average value > 10.0. The slopes for probit models calculated for the rest of the tested insecticides ranged from 1.00 to 10.0 (S1-S4 Tables).
Extensive surveys have shown that stronger resistance to neonicotinoids has been confirmed in some populations of Bemisia tabaci and Leptinotarsa decemlineata [13], Triaulerodes vaporarium [14], Nilaparvata lugens [15] and Myzus persicae [16]. A study of L. decemlineata resistance to imidacloprid in populations from North America and Europe revealed an approximately 30-fold variation in the LC50 values from ingestion and contact bioassays against neonates [17]. After several years of imidacloprid application, 100-fold [18] and 309-fold [19] resistance ratios for imidacloprid were reported in populations of L. decemlineata from Long Island.
No shift to lower susceptibility to thiacloprid has been observed in populations of B. aeneus and C. obstrictus from several European countries in glass vial experiment with a commercially available thiacloprid formulation, Biscaya OD 240 [20]. According to Zimmer and Nauen [21], the adult vial test with an oil-dispersion-based formulation of thiacloprid resulted in a much better bioavailability compared with technical material.
Resistance to neonicotinoids has been slow to develop compared with other insecticide groups [13]. A target-site resistance has been documented from a selected neonicotinoid resistant laboratory strain of N. lugens [22]. However, this mutation has not been found in resistant strains of N. lugens collected from the field [23]. From recent evidence, it is clear that neonicotinoid resistance is at present based largely on P450-dependent monooxygenases [24]. It was found that up-regulation of CYP6CM1 in B. tabaci confers resistance not only to several neonicotinoids but also to pymetrozine [25]. Puinean et al. [16] published evidence that a P450 gene amplification event is associated with neonicotinoid resistance in M. persicae. Recently, R81T mutation conferring target-site resistance to neonicotinoids was recorded in M. persicae from southern Europe [26].
No cross-resistance between thiacloprid and acetamiprid or thiamethoxam was recorded for CSFB. Cross-resistance to neonicotinoids was described for several populations and biotypes of the whitefly (B. tabaci) [13]. Cross-resistance between imidacloprid and thiamethoxam and between acetamiprid and thiamethoxam was observed in B. tabaci [27]. Crossresistance between imidacloprid, thiacloprid and acetamiprid was detected in N. lugens [1]. Similarly, cross-resistance between imidacloprid and acetamiprid was detected in M. persicae [28]. In the present study, the tolerance of CFSB to thiacloprid was detected with preserved susceptibility to acetamiprid and thiamethoxam. Tolerance to thiacloprid was also demonstrated in honeybees and bumble bees [29]. It was found, that the sensitivity of Apis mellifera and Bombus terrestris to neonicotinoids is determined by cytochrome P450s of the CYP9Q subfamily and that CYP9Q3 metabolizes thiacloprid with high efficiency but has little activity against imidacloprid. Further research is necessary to prove whether similar mechanism could explain the tolerance of CSFB to thiacloprid and sensitivity to thiamethoxam. Little or no cross-resistance of neonicotinoids to older insecticide classes, such as pyrethroids, organophosphates and carbamates, has been observed [30]. One strain of the small brown planthopper (Laodelphax striatellus) maintained under laboratory conditions apparently acquired an 18-fold resistance to imidacloprid, without exposure to this insecticide, under strong selection from organophosphates and carbamates, while the unselected strains were fully susceptible to imidacloprid [31].

Susceptibility to neonicotinoids in the feeding experiment
The feeding experiment showed different susceptibilities of the Prague CSFB population to thiacloprid and thiamethoxam. The samples were highly susceptible to thiamethoxam and tolerant to thiacloprid. The injury to plants in the thiacloprid variant was comparable to the control variant without seed dressing in the feeding experiment. In contrast to this, the seed dressing with thiamethoxam importantly prevented CSFB injury in oilseed rape plants ( Table 2). The increase in the average weight of plants in the thiamethoxam variant was observed as a result of the prevention of CSFB injury. A very low susceptibility of CSFB to thiacloprid and also to acetamiprid was reported by Dewar [32].
The experiments proved that seed dressing with thiamethoxam was highly effective for the control of oilseed rape plants in their early growth stages against CSFB. As a consequence of the restriction of seed dressing with three neonicotinoids by the European Commission [5], the injury to oilseed rape plants caused by CSFB increased. It resulted in the increased frequency of the foliar application of insecticides and increased costs for the control of oilseed rape in autumn. The increase in the foliar application of insecticides was also documented in the western part of the EU [6]. Seed dressing with thiacloprid had a very low efficacy due to the CSFB tolerance.

Conclusions
The study showed good efficacy of the pyrethroids lambda-cyhalothrin, deltamethrin, cypermethrin, tau-fluvalinate, esfenvalerate and etofenprox, organophosphate chlorpyrifos and oxadiazine indoxacarb against two local CSFB populations in the Czech Republic in 2015-2018. Both the CSFB populations were susceptible to nenonicotinoids acetamiprid and thimethoxam and tolerant to thiacloprid. The CSFB tolerance to thiacloprid was shown in both glass vial and feeding experiments.
Supporting information S1