Resistance of Spanish codling moth (Cydia pomonella) populations to insecticides and activity of detoxifying enzymatic systems

Resistance of codling moth, Cydia pomonella (L.) (Lepidoptera: Tortricidae), to insecticides has become a major problem in many apple and pear production areas. Our aim was to determine the level of insecticide resistance in Spanish field populations. Seven field populations collected from apple, Malus domestica Borkhausen (Rosaceae), orchards, and three laboratory susceptible strains of codling moth were studied. Damage at harvest in all the conventional orchards from which codling moth populations were collected was higher than the economic threshold. The efficacy of eight insecticides, with five modes of action, was evaluated by topical application of the diagnostic concentrations on post‐diapausing larvae. The enzymatic activity of mixed‐function oxidases (MFOs), glutathione transferases (GSTs), and esterases (ESTs) was evaluated for each population. The susceptibility to insecticides and the biochemical activity of the three laboratory strains and one organic orchard population were not significantly different. Field populations were less susceptible to the tested insecticides than the susceptible strains, especially for azinphos‐methyl, diflubenzuron, fenoxycarb, and phosalone. The efficacy of all insecticides was significantly dependent on the activity of MFOs. Only the toxicity of the three insecticides most used in Spain when the populations were collected (azinphos‐methyl, fenoxycarb, and phosalone) was also dependent on the activity of ESTs and GSTs activity. We conclude that the control failures were because of the existence of populations resistant to the main insecticides used.


Introduction 44
The codling moth, Cydia pomonella (L.) (Lepidoptera: Tortricidae), is a worldwide key pest 45 in apple, pear, quince, and walnut orchards in temperate regions (Shel'Deshova 1967). The 46 geographic origin of this pest is Eurasia, from where it has spread to European regions, North 47

Insecticide efficacy 145
The efficacy of each insecticide was tested at the diagnostic concentration shown in Table 1. 146 The tested diagnostic concentrations have been reported to cause 72-99.9% mortality in a 147 susceptible Swiss population (S_Switzerland) (Pasquier & Charmillot, 2003;Charmillot et al., 148 2007). A 1-µl drop of the insecticidal solution -or of the organic solvent in the case of the 149 controls -was applied with a Multipette Plus (Eppendorf) in the dorsal median region of the 150 larva (Sauphanor et al., 2000). Immediately after the treatment, 10 larvae were transferred 151 inside 20 × 20 mm pieces of corrugated cardboard placed within 90-mm-diameter Petri 152 dishes, and were kept at 22 ± 3 ºC and L16:D8. The Petri dishes were checked daily until 15 153 days had elapsed since the last adult emergence, and the mortality was then recorded. Four 154 replicates of 10 larvae each were carried out per population and insecticide. Depending on the 155 sample size, 3-8 insecticides were tested per field population. 156 157

Enzymatic activity 158
Twenty post-diapausing larvae were used per population and enzymatic complex. Mixed-159 function oxidase (MFO) activity was measured by fluorescence, and glutathione transferase 160 (GST) and esterase (EST) activity was measured by absorbance (Bouvier et al., 2002), using a 161 VICTOR 3 Multilabel Plate Counter (PerkinElmer Life and Analytical Sciences, Madrid, 162 Spain). 163 The MFO activity was analyzed with an in vivo protocol. Post-diapausing larvae were 164 individually placed on ice in a 6 g l -1 sodium chloride solution, and were cut, using micro-165 scissors, into four fragments in order to prevent over-expression of the enzymatic activity per 166 well. Each fragment was deposited individually in a well and used as the enzymatic source. 167 The results of the four fragments were added to obtain the enzymatic activity per individual. 168 The GST and EST activities were also analyzed with an in vivo protocol. To obtain the 169 enzyme extracts, post-diapausing larvae were individually homogenized on ice in 1 ml of a 170 solution of phenyl-methylsulfonyl fluoride (PMSF, 0.4 mM) on phosphate buffer (50 mM, pH 171 7.2). The homogenates were centrifuged for 15 min at 4 ºC and 15 000 g, and the supernatant 172 of each sample was divided into two aliquots that were used as the enzymatic sources of GST 173 and EST (Bouvier et al., 2002). 174

175
Determination of MFO activity 176 The MFO activity was determined in black PORVAIRpic 96-well microplates (BIOGEN 177 Científica, Madrid, Spain) using the 7-ethoxycoumarin-O-deethylation (ECOD) method 7 (Bouvier et al., 2002). The microplates were maintained on ice. Each fragment of the post-179 diapausing larvae was placed individually in a well with 100 µl of sodium phosphate buffer 180 (pH 7.2, 50 mM) and 7-ethoxycoumarin (0.4 mM). As controls, 12 wells were supplied with 181 100 µl of glycine/ethanol buffer (vol/vol; 0.1 mM, pH 10.4) immediately after the addition of 182 the larval tissue, in order to avoid the reaction. After 4 h of incubation at 30 ºC, the reaction 183 was stopped by adding 100 µl of the glycine/ethanol buffer and the microplate was 184 centrifuged at 2 000 g for 1 min in order to immerse the larvae fragment and clear the surface 185 of the well for the reading. were compared with that of S_Spain by a χ 2 test. When significant differences were detected, 227 an enzymatic activity ratio was calculated by dividing the value of the enzymatic activity of 228 each population by the value of S_Spain. When no significant differences were detected, the 229 value of the enzymatic activity ratio was 1. When the number of tested populations was 230 greater than nine, the relationship between each insecticide efficacy and each enzymatic 231 activity was analyzed by linear regression analysis, using enzymatic activity as the 232 independent variable. When the number of tested populations was greater than seven, the 233 relationship among the efficacies of each insecticide and among the enzymatic activities was 234 analyzed by correlation analysis. 235 236

Enzymatic activity 256
Activity of MFOs, GSTs, and ESTs in C. pomonella post-diapausing larvae was dependent on 257 the population (F9,190 = 7.36, 8.53, and 9.37, respectively, all P<0.0001 ), but no significant 258 differences were observed among the three susceptible strains and the population from the 259 organic orchard (Table 3). Most of the other field populations (either from the mating 260 disruption + insecticide orchard or from the conventional orchards) showed a significantly 261 higher MFO, GST, and EST activity than that of S_Spain (Table 3). The mating disruption + 262 insecticide orchard population showed MFO and EST activity levels similar to those of the 263 conventional orchard populations mentioned above (Table 3). All the field populations 264 collected from the conventional orchards showed a significantly higher MFO and GST 265 activity than S_Spain, whereas four out of five field populations from conventional orchards 266 showed higher EST activity than S_Spain (Table 3). The values of the enzymatic activity ratio 267 were much higher for MFO activity (ranging from 8.4 to 74.6) than for GST (1.6-2.6) and 268 EST activity (3.3-7.0) ( Table 3). 269 The thresholds used to determine the frequency of resistant individuals in the field 270 populations were 14.14 pg 7OH per larva per min, 13.7 mM of glutathione conjugated mg per 271 protein per min, and 451.39 nmol of β-naphtol mg per protein per min for RMFO, RGST, and 272 REST, respectively. The RMFO, RGST, and REST of S_France, S_Italy, and the organic 273 orchard population were not significantly different from the those of S_Spain, whereas the 274 rest of the field populations showed a higher frequency of resistant individuals than S_Spain, 275 except in two cases (Table 3). For the populations from non-organic orchards, the RMFO 276 ranged from 95 to 100%, the RGST from 20 to 80%, and the REST from 20 to 100%. 277 278 279

Relationships between insecticide efficacy and enzymatic activity 280
The insecticide efficacy was significantly dependent on MFO activity for all insecticides 281 tested (Table 4,  GST activities, respectively) ( Table 4, gray area). 287 The efficacies of the insecticides were positively intercorrelated in most cases (Table 4,  288 white area). The efficacy of azinphos-methyl, fenoxycarb, and tebufenozide was significantly 289 correlated with that of the other products tested, except in the case of chlorpyriphos-methyl. 290 On the other hand, the efficacy of phosalone and diflubenzuron showed no significant 291 correlation with that of the neonicotinoid thiacloprid, and the efficacy of chlorpyrifos-ethyl 292 was not correlated with that of chlorpyriphos-methyl and thiacloprid (Table 4, white area). A 293 significant correlation was found between the efficacy of chlorpyriphos-methyl and that of 294 only two other insecticides: diflubenzuron and phosalone. Similarly, the efficacy of 295 thiacloprid was significantly correlated only with that of azinphos-methyl, fenoxycarb, and 296 tebufenozide (Table 4, white area). The EST activity was positively correlated with that of 297 MFO and GST, but the MFO and GST activities were not significantly correlated (Table 4,  298 white area). 299 300

Discussion 301
The susceptibility of the three susceptible strains, S_Spain, S_France, and S_Italy, to the eight 302 insecticides tested may be considered essentially equal, as the efficacy of insecticides on them 303 was significantly different only in two cases out of 16. (Table 2). The efficacies of the 304 different insecticides were very similar to those reported for the susceptible strain 305 S_Switzerland (Pasquier & Charmillot, 2003). Similarly, the MFO, GST, and EST activity of 306 the three susceptible strains was equal (Table 3). The importance of these results lies in the 307 use of a particular susceptible strain to calculate the resistance rate of field populations. As the 308 three tested European susceptible strains are equal, the resistance ratios of field populations 309 from different countries may be compared. Though Silva et al. (2003) found that the 310 geographic origin and the laboratory rearing period could produce differences between 311 susceptible strains, in our case these factors did not influence the results. After these results, 312 S_Spain was selected as the reference susceptible strain for present and further studies. 313 11 Only the population from the organic orchard, Boldú, may be considered as a 314 susceptible field population, because the efficacy of all tested insecticides on it was not 315 significantly different from that on S_Spain, with the exception of fenoxycarb. This 316 population showed an enzymatic activity that was not significantly different from that of 317 S_Spain for the three enzymatic systems. In contrast, the population collected from the mating 318 disruption + insecticide orchard (Gimenells, which has a low use of insecticides) showed low 319 susceptibility to all the insecticides tested, except thiacloprid, and higher MFO and EST 320 activities than those of S_Spain, probably because the Gimenells orchard is surrounded by 321 conventional orchards, whereas the Boldú orchard has no other apple or pear fruit orchards in 322 its vicinity. The migration of codling moths among orchards with different management 323 programs has been recorded in several countries (Knight et al., 1994;Fuentes-Contreras et al., 324 2007), where insecticide resistance has been reported even in populations collected from 325 abandoned orchards due to their proximity to commercial orchards. 326 All the field populations from conventional orchards were resistant to azinphos-methyl, 327 phosalone, diflubenzuron, and fenoxycarb, with considerable decreases in the insecticide 328 efficacy, which was in most cases lower than 50%. Chlorpyrifos-methyl and chlorpyrifos-329 ethyl were more toxic to codling moth than phosalone and azinphos-methyl, as has already 330 been reported for European (Reyes et al., 2007;Rodríguez et al., 2010) and North American 331 (Dunley & Welter, 2000) codling moth populations. Azinphos-methyl and phosalone (which 332 have been prohibited in Spain only since 2006 and 2007, respectively) have been much more 333 widely used in Spain than chlorpyrifos-methyl and chlorpyrifos-ethyl. Tebufenozide has been 334 used little in Spain, but it showed low efficacy on field populations. As we found a positive 335 correlation between the efficacy of tebufenozide and the efficacy of six insecticides, including 336 the organophosphate azinphos-methyl and the IGR fenoxycarb, the low efficacy of 337 tebufenozide may be attributed to cross-resistance, as has been reported several times for 338 codling moth (Ioriatti et al., 2007;Knight et al., 2001;Reyes et al., 2007;Sauphanor & 339 Bouvier, 1995) and other lepidopterous pests (Smirle et al., 2002). 340 The field populations in this study were in general more susceptible to thiacloprid -not 341 yet used in Spain when the populations were collected -than to the other insecticides, 342 although the susceptibility of the conventional orchard populations was slightly lower than 343 that of the susceptible strains. Deciding whether this fact points to an incipient resistance due 344 to cross-resistance needs further investigation with more populations, as we also found a 345 positive correlation among the efficacies of thiacloprid, azinphos-methyl, fenoxycarb, and 346 tebufenozide, and cross-resistance has been reported (Reyes et al., 2007). 347    Table 4 Matrix of correlation coefficients (white area) and determination coefficients (gray area) of the correlation analyses and of the linear 1 regression analyses of insecticide efficacy on enzymatic activity for three susceptible strains and seven field populations of Cydia pomonella 2 Insecticides: see Table 1 for full names. Enzymes: MFO, mixed-function oxidases; GST, glutathione transferases; and EST, esterases. 4