Chemical
Extraction of E. nitens essential oil and its characterization by GC-MS was described by Alvarez Costa et al. (2017). N,N-diethyl-m-toluamide DEET (97%) and 1,8-cineole (99%) were purchased from Sigma-Aldrich® (St. Louis, MO). Acetone (pro-analysis, 99.5%) and hexane (pro-analysis, 99.5%) were purchased from Sintorgan®, Argentina.
Biological material
Ae. aegypti colony is reared since 1996 in our insectarium, and it was derived from the Rockefeller strain from Venezuela. These mosquitoes are reared at 25 ± 2°C, 70-80% RH, and a photoperiod of 12:12 h (L:D). The colony is susceptible to insecticides and have been free of exposure to pathogens or repellents. A more detailed protocol of the rearing conditions is described in Gonzalez et al. (2016).
An. pseudopunctipennis immature stages were collected in the field from natural breeding sites, and transported to a nearby laboratory in San Ramón de la Nueva Oran, Salta, Argentina (23°08′10″S 64°19′20″W). Larvae were maintained in dechlorinated water with a portion of the filamentous algae (Spyrogira sp.) from their breeding sites. Adult mosquitoes were fed only on raisings. At the end of each behavioral assay, the organisms were identified with taxonomic keys (Darsie 1985).
Ethics approval and consent to participate
No human participants, human data or human tissue were used in the present study. Insects were fed on pigeon blood once a week according to a protocol approved by the Institutional Animal Care and Use Committee of CIPEIN (IACUC/CICUAL 1531/13).
Plaque repellency assay
The repellent activity test was based on the Bar-Zeev (1960) protocol with some modifications. A half filter paper (9 cm total diameter) was impregnated with 0.5 mL of the testing substance at different concentrations, and the other half was impregnated with solvent (acetone). In control group both halves of the filter paper were impregnated with acetone. The solvent was left to evaporate for an hour and then the paper was placed in a Petri dish (9 cm of diameter and 3 cm of height). Females were placed individually in the Petri dish and quickly closed to avoid them to escape. Each female was digitally recorded for 10 min with a video-camera (Panasonic Lumix DMS-LS 80). The humidity and ambient temperature were maintained at 25 ± 2°C and 70-80% RH respectively. To ensure enough contrast between the female and the background a light under the Petri dish (experimental arena) was used. Five to ten days non-blood-fed females were used for the assays. Ten replicates for An. pseudopunctipennis and 15 for Ae. aegypti were performed for each substance and concentration. The substances evaluated were DEET, E. nitens essential oil and their major component, 1,8-cineole (22.9%) (Alvarez Costa et al. 2017).
We used the tracking software (EthoVision XT10.1) to analyze the recorded videos and to calculate behavioral variables of the females (Noldus et al. 2001). To identify the adult from their background we used dynamic subtraction. The rate at which the program analyzes the video to find the insect, the sample rate, was three samples per second.
To evaluate the repellent effect, the total time (in seconds) that the mosquitoes remained at the control and the treated zone were obtained. Then a repellence index (RI) was calculated, using the following formula:
This index varies from 0 to 1, being 0 when the mosquito spent all of the time in the control zone and 1 when it spent all of the time on the treated zone. A RI of 0.5 means that the mosquito spent half time on each zone.
In addition, to evaluate how the exposure to repellent substances would affect the locomotor activity of the females, we studied the following locomotor variables: distance travelled by the female, its median velocity, its absolute angular velocity (AAV) and its mobility state, the cumulative time of a discrete state variable with three possible states: highly mobile (H), mobile (M) or immobile (I) (Noldus et al. 2001; Alvarez Costa et al. 2018).
Statistical analysis
To evaluate the repellent effect ANOVA of each substance and species were performed using RI as the response variable. To evaluate the locomotor activity of the females, first, the information of the activity variables (velocity, distance, AAV and the time in each mobility state) was condensed performing a Principal Component Analyses (PCA). The mean and standard error of each activity variable for each experiment is shown in the Supplementary Table 1. A PCA was made for each substance and species. PCA generates abstracts axis (principal components), the biological meaning for each PC 1 depends on their correlations with the original variables. Then, to characterize the PC 1, the association between the axe and the variables obtained was quantified with Spearman correlations. All PC 1s obtained were assigned as an ‘activity axis‘. For both ANOVA the LSD Fischer a posteriori test were performed to compare the differences between means of the treatments. The assumptions were tested graphically and by Levene and Shapiro Wilks tests. We used the Bonferroni correction to maintain a global error of 0.05, then the threshold for significance was set at P< 0.025.