How varying parameters impact insecticide resistance bioassay: An example on the worldwide invasive pest Drosophila suzukii

Abstract

Monitoring pesticide resistance is essential for effective and sustainable agricultural practices. Bioassays are the basis for pesticide-resistance testing, but devising a reliable and reproducible method can be challenging because these tests are carried out on living organisms. Here, we investigated five critical parameters and how they affected the evaluation of resistance to the organophosphate phosmet or the pyrethroid lambda-cyhalothrin using a tarsal-contact protocol on Drosophila suzukii, a worldwide invasive pest. Three of the parameters were related to insect biology: (i) sex, (ii) age of the imago (adult stage) and (iii) genetic diversity of the tested population. The two remaining parameters were linked to the experimental setup: (iv) the number of individuals tested per dose and (v) the duration of exposure to the active ingredient. Results showed that response to insecticide differed depending on sex, males being twice as susceptible to phosmet as females. Age principally affected young females' susceptibility to phosmet, because 0-24 hour-old flies were twice as susceptible as 24-48 hour-old and 72-96 hour-old females. Genetic diversity had no observable effect on resistance levels. The precision and accuracy of the median lethal dose (LD50) were greatly affected by the number of individuals tested per dose with a threshold effect. Finally, optimal duration of exposure to the active ingredient was 24 h, as we found an underestimation of mortality when assessed between 1 and 5 h after exposure to lambda-cyhalothrin. None of the main known point mutations on the para sodium channel gene associated with a knockdown effect were observed. Our study demonstrates the importance of calibrating the various parameters of a bioassay to develop a reliable method. It also provides a valuable and transferable protocol for monitoring D. suzukii resistance worldwide.

Fig 1. Effect of sex on LD₅₀ in a D. suzukii population.

Dose-response curves of 24 to 48 h old male (light gray triangles, shading and dashed line) and female (dark gray circles, shading and solid line) adults of D. suzukii after 24 h of tarsal exposure to phosmet. The 95% confidence intervals were derived from the dose-response model.

Fig 2. Effect of age class on LD₅₀ in a D. suzukii population.

Results obtained after 24 h tarsal exposure to phosmet for females (upper panel) and males (lower panel). Red, green and blue represent the 0–24 h, 24–48 h and 72–96 h age classes, respectively. Dose-response curves and 95% confidence intervals were derived from the dose-response model.

Fig 3. Diversity indices of the Ste Foy and SF-IsoA populations.

Radar plot of (A) gene diversity (H_e) and (B) the number of observed alleles (N_a) at 13 microsatellite loci for the Ste-Foy (red) and SF-IsoA populations (blue).

Fig 4. Effect of population genetic diversity on LD₅₀ in D. suzukii.

Dose-response curves of D. suzukii 24 to 48 hour-old adults of the population Ste-Foy (blue) and the inbred SF-IsoA (red) derived from it, after 24 h of tarsal exposure to phosmet in females (top panel) and males (bottom panel). The 95% confidence intervals were derived from the dose-response model.

Fig 5. Effect of the mean number of individuals tested per dose on LD₅₀ in a D. suzukii population.

Distribution of LD₅₀ values and associated 95% confidence intervals after 24 h of tarsal exposure of 24 to 48 hour-old adults to phosmet, according to the number of individuals tested for females (left panel) and males (right panel). The continuous red line indicates the LD₅₀ value computed on pooled individuals as one unique bioassay and the dashed lines show the corresponding 95% confidence interval.

Fig 6. Effect of duration of exposure on LD₅₀ estimates in a D. suzukii population.

The bioassay consisted of a tarsal exposure to lambda-cyhalothrin and was conducted on 24 to 48 hour-old D. suzukii adults from the Ste-Foy population. Panel A shows the change in the dose-response curves for female individuals. Early scoring (after 1, 2, 3, 4 and 5 h) curves are shown in black and late scoring (after 20, 21, 22, 23 and 24 h) curves are shown in red. The black arrow shows the variation in LD₅₀ between 1 and 5 h. The half-matrix in Panel B shows the significance levels of relative potencies between different exposure times for the female individuals (red: p< 0.001; orange: 0.001 ≤p <0.01; yellow: 0.01 ≤ p <0.05; gray: ns).