Common facultative endosymbionts do not influence sensitivity of cereal aphids to pyrethroids

Cereal aphids, including the bird cherry-oat aphid, Rhopalosiphum padi, and the grain aphid, Sitobion avenae, can transmit viruses that significantly reduce crop yields. To mitigate against yield losses, insecticides are routinely used to manage aphid populations. Aphids can form relationships with endosymbionts that confer fitness benefits or consequences to the aphid. Recent artificial inoculation experiments indicate that endosymbionts could increase aphid susceptibility to insecticides, but this has not been explored using aphid populations naturally infected with endosymbionts. Here, we sampled aphids from an important cereal production region in Lower Saxony, Germany. We characterised the endosymbiont profile of these aphid populations and conducted pyrethroid dose-response assays to test the hypothesis that facultative endosymbionts increase aphid susceptibility to insecticides. We find that the level of insecticide susceptibility is highly variable in S. avenae and we identify populations that are sensitive and tolerant to pyrethroids, including populations collected from the same field. For R. padi, we find evidence for decreased sensitivity to pyrethroids, representing the first report of reduced sensitivity to pyrethroids in R. padi sampled from Central Europe. We detected high endosymbiont infection frequencies in the aphid populations. 84% of aphids carry one facultative endosymbiont and 9% of aphids carry two facultative endosymbionts. We detected associations with Regiella insecticola, Fukatsia symbiotica, and Hamiltonella defensa. However, we do not identify a link between endosymbiont infection and insecticide susceptibility, indicating that other factors may govern the development of insecticide resistance and the need for alternative management strategies.

4. We find that the level of insecticide susceptibility is highly variable in S. avenae and 21 we identify populations that are sensitive and tolerant to pyrethroids, including 22 populations collected from the same field. For R. padi, we find evidence for 23 decreased sensitivity to pyrethroids, representing the first report of reduced 24 sensitivity to pyrethroids in R. padi sampled from Central Europe. 25 5. We detected high endosymbiont infection frequencies in the aphid populations. 84%

26
of aphids carry one facultative endosymbiont and 9% of aphids carry two facultative 27 endosymbionts. We detected associations with Regiella insecticola, Fukatsia 28 symbiotica, and Hamiltonella defensa. However, we do not identify a link between 29 endosymbiont infection and insecticide susceptibility, indicating that other factors 30 may govern the development of insecticide resistance and the need for alternative 31 management strategies. 40 (Nancarrow et al., 2021).

41
Insecticides remain the method that is most commonly used to manage aphid populations, with 42 pyrethroids widely used for the management of cereal aphids on spring and winter cereal crops 43 across Europe (Dewar & Foster, 2017). The high reliance on pyrethroid insecticides increases the 44 evolutionary pressure on aphid populations, increasing the risk that insecticide resistant aphid 45 populations will emerge (Dewar & Foster, 2017). Insecticide resistant populations can have 46 devastating consequences on effective aphid management and increase potential aphid-derived 47 yield loss (Dewar & Foster, 2017), making these an urgent priority for the development of

74
The lack of high prevalence of resistant populations across regions and years (Walsh et al., 2020b; 75 Gong et al., 2021)

115
Here, we report the results of pyrethroid dose-response bio-assays for cereal aphid populations 116 sampled from a key cereal production region in Northern Germany. We sampled 25 S. avenae and 117 seven R. padi populations from 13 field sites. We find that, for S. avenae, the level of insecticide 118 susceptibility is highly variable, with populations sensitive and tolerant to pyrethroid exposure, 119 including populations collected from the same field. In R. padi populations, we find evidence for 120 decreased sensitivity, indicating that resistance to pyrethroids is starting to evolve in German R.

Insecticide sensitivity testing 138
Aphid populations were screened for susceptibility and sensitivity to the synthetic pyrethroid Decis 139 Forte® (Bayer CropScience, Germany), a Class 3A synthetic pyrethroid (active ingredient 140 deltamethrin at 100 g L -1 formulation). This insecticide was selected as it is approved for use on 141 arable and field crops in Germany. A stock solution was prepared in water at a concentration 142 comparable to the recommended field rate, equating to a concentration of 357 mg a.i. L -1 . Serial 143 dilutions were prepared from the stock solution. Five insecticide dilutions were used in the assay: 144 stock, 10 -1 , 10 -2 , 10 -3 , 10 -4 , with distilled water included as a negative control.

145
The insecticide sensitivity assays broadly followed the IRAC leaf-dip method (IRAC, 2016) and the 146 method deployed by (Umina et al., 2020). Briefly, c. 25 mm sections of wheat leaves were 147 submerged for c. 10 s in one of the test solutions. Control leaves were dipped first, then the leaves 148 5 were dipped sequentially from the lowest concentration (10 -4 ) to the field rate stock solution. Once 149 dipped, leaves were left to dry on paper towels for approximately 1 h before they were placed 150 abaxial side up on agar (1 g L -1 ) in a plastic Petri Dish; a droplet of water was added to the surface 151 of the agar to aid leaf adhesion. Aphids were transferred to each Petri Dish using a fine-haired 152 paintbrush, Petri Dishes were moved to a controlled environment room (20°C ± 2°C, L16:D8), and 153 Petri Dishes were inverted to simulate aphid feeding from the underside of the leaf. Between 4 -8 154 aphids were transferred to each Petri Dish. After 48 h aphids were scored as either alive, 155 moribund, or dead. Aphids were classed as alive if they were able to return to an upright position 156 when placed on their back (i.e., they were capable of coordinated movement). Moribund and dead 157 aphids were grouped together as "affected", in-line with previous dose-response assays (Foster et

DNA extraction and diagnostic PCR for endosymbiont characterisation 160
A sample of five aphids (mixture of apterous adults and nymphs) were collected from each  and Arsenophonus spp. All PCR primer details are described in Table S1. PCR assays were  Table S1. The final reaction mixture was made to 12 µL using nuclease-

Statistical analysis 185
All statistical analysis was carried out using R (v.4.1.2) and R Studio (v.1.3.1093). The estimated 186 concentration of active ingredient required to achieve 50% mortality (EC 50 value), EC 50 95% 187 confidence intervals, slope, intercept, and associated standard errors for the dose response curve 188 were calculated for each aphid population using probit estimation regression (Finney, 1952). To 189 achieve this, the "ProbitEPA" function in the R package ecotoxicology (v.1.0.1) was used.  Simpson's diversity was calculated using the vegan package (v.2.5-7). Where significant 196 differences in model intercepts were detected, the differing aphid populations were identified by 197 observing the overlapping confidence intervals. This method has been used previously to identify 198 differences in insecticide susceptibility between aphid populations from estimated concentration of 199 active ingredient required to achieve 50% population mortality (EC 50 ) values (Foster et al., 2012).

Pyrethroid sensitivity is variable in Sitobion avenae populations 202
Based on mortality at field rate concentration (i.e., mortality in the stock treatment, 357 mg a.i. L -1 203 deltamethrin), S. avenae populations were grouped into four broad categories (Table 1) Table 1).

211
The estimated effective dose required for 50% population control (EC 50 ) ranged from 0.38 mg a.i.

236
The estimated effective dose required for 50% population control (EC 50 ) ranged from 0.44 mg a.i.

Facultative endosymbionts occur at high frequencies in aphid populations but they 247 do not influence pyrethroid sensitivity 248
Of the 32 aphid populations 27 were infected with at least one facultative endosymbiont (Table 2).

249
The endosymbiont community differed between the two cereal aphid species: endosymbionts were 250 detected in 92% of S. avenae populations and 57% of R. padi populations (

369
One caveat of our study was our lack of a characterised kdr-SS homozygous pyrethroid

Data accessibility 388
The data that support the findings of this study are available from the corresponding author upon 389 reasonable request.