Spider mites were provided by Vineland Research and Innovation Centre (Vineland Station, ON, Canada) to establish a colony at Acadia University, Wolfville (NS, Canada). All mites used in this study were adult females: they are easily identifiable by their large size and blunt abdomen relative to males (Meena et al., 2013). The colony was maintained on common bean plants (Phaseolus vulgaris L.) in wire mesh cages (100 x 100 x 100 cm) located in a growth chamber (25 ± 2°C, 16:8 L:D, 70 ± 5 % RH). Bean plants were changed every 2 days, replacing several of the oldest plants with 3-week-old plants. Bean plants were grown in phytotrons in the Harriet Irving Botanical Gardens (K.C. Irving Environmental Science Centre, Acadia University, NS, Canada), kept between 18 and 25°C on a 12:12 light/dark photoperiod.

A variety of essential oil components, organic acids and aldehydes were chosen for testing. The chemicals and their sources include: alpha-phellandrene, linalool (97%) (+/-) racemic mixture, nonanal (95%), thymol (≥98%), eucalyptol (99%), propionic acid (≥99.5%, FCC, FG), isovaleric acid (≥99%), hexanoic acid (≥99%), octanoic acid (>99%), pentanal (≥99%), hexanal (98%), and octanal (≥99%), which were all purchased from Sigma-Aldrich (Saint Louis, MO, USA). Butyric acid (≥98%), Cis-3-hexen-1-ol (>99%), and Cis-3-hexenyl-acetate (>99%) were bought from Bedoukian (Danbury, CT, USA). Beta-phellandrene (80%) was purchased from Foreverest (Xiamen, Fujian, China). Farnesene (mixture of isomers) and heptanal (92%) were boughtfrom SAFC (Steinheim, Germany). HPLC-grade hexane was purchased from Fisher Scientific (Fair Lawn, NJ, USA). Eucalyptus, coriander essential oil, and neem oil were bought from Ultra International B.V. (LA Spijkenisse, Netherlands). Lavender essential oil was purchased from New Directions Aromatics (Mississauga, ON, Canada). Freshly extracted basil essential oil was provided by Dr. Faraone (Chemistry Department, Acadia University). Tween-80 was purchased from Sigma-Aldrich (Saint Louis, MO, USA).

A novel electrophysiological approach was developed to record olfactory sensitivity of female T. urticae using a Syntech EAG fork electrode (Ockenfels Syntech, Buchenbach, Germany). A small amount of electrode gel (approximately 0.1mL) (Signagel^®, Parker Laboratories Inc, NJ, United States) was placed on the base electrode, and a single female mite was mounted dorsally within the gel, with her gnathosoma facing the recording electrode ( Figure 1 ). The foretarsi were attached underneath the recording electrode with additional electrode gel. Then the tips of the foretarsi were cut with a scalpel (Sigma-Aldrich, Saint Louis, MO, USA) to help establish a connection for the electrotarsography recordings. Given the small size of T. urticae (an adult female is approximately 0.5mm), all mounting preparations required the use of a dissecting microscope (AmScope SM-1BSX-64S, Irvine, CA, USA).

Electrotarsography (ETG) was used to screen our selected compounds to which T. urticae can sense. We focused on screening compounds known to be detected by ionotropic rectors, such as carboxylic acids and aldehydes (Ni, 2021), and various active ingredients of essential oils known to be potentially effective for pest management. All ETG and GC-ETD recordings were amplified via the Intelligent Data Acquisition Controller-2 (IDAC-2) (Syntech, Kirchzarten, Germany) and recorded through Data acquisition for Gas Chromatograph with EAD (GcEad 2012 v1.2.4 (2012-06-24), Syntech, Kirchzarten, Germany).

ETG measured the stimulation of the olfactory sensilla of a female mite specimen to a given compound within a stimulus cartridge when a puff of air is directed through the cartridge to the mounted specimen. The voltage recorded in response to specific stimuli corresponded to a specimen’s sensitivity to a particular compound.

Stimulus cartridges contained a single piece of filter paper (7 mm x 35 mm Fisherbrand, Fisher Scientific, ON, Canada) loaded with 10 μL of the compound of interest. The loaded filter paper was placed into a disposable borosilicate glass pipette (Fisherbrand, Fisher Scientific, ON, Canada). Then the glass pipette was capped with a 1-mL disposable pipette tip (Fisherbrand, Fisher Scientific, ON, Canada) and sealed with dental wax (Electron Microscopy Sciences, Hatfield, PA, USA). The following compounds were tested: α-phellandrene, β-phellandrene, butyric acid, farnesene, linalool, (Z)-3-hexen-1-ol, (Z)-3- hexen-1-yl acetate, and nonanal. Compounds of interest were diluted in HPLC-grade hexane and tested at 10 ng, 100 ng, 1 μg, and 10 μg. All prepared stimulus cartridges were stored at -20°C until use. Stimulus cartridges were reloaded with diluted compounds after 1 month or a maximum of 20 puffs based on preliminary data.

All compounds were tested in randomized order with increasing dosages (10 ng to 10 μg). A single odorant puff lasted 0.3 s controlled by a Syntech stimulus controller CS-55 V2.7 (Ockenfels Syntech). The flow rate of air for each puff is 3.54 L/min. Each stimulus cartridge was puffed in intervals of 30 s to allow for preparation recovery (time to recover from last stimulus and be prepared for next stimulus).

After screening both essential oil active ingredients, carboxylic acids, and aldehydes with ETG, three different master mixes were prepared based on combination of essential oils, carboxylic acids, or aldehydes. We test master mixes (e.g., selected individual compounds all dissolved in the same vial using the same solvent) to minimize the number of spider mite preps required for the acquisition of the recordings. The different master mixes were screened using the Varian 450-GC with a flame ionization detector (GC-FID) and a non-polar DB-5 Column (30m, 0.25mm, 0.25μm film thickness) (Varian Inc., Lake Forest, CA, USA) linked to an electrotarsal detector (GC-ETD), to detect responses to individual components of the mixtures. Each compound added to the essential oil master mix had a final concentration of 100 ng/μL. However, it is important to recognize that the individual compounds of each essential oil mixture vary in concentration, and that some compounds exist in more than one essential oil mixture, which increases the overall concentration. This may affect the recorded peak amplitude based on a dose-response relationship. Since the objective of the electrotarsogram study was only to identify through a qualitative analysis which compounds the mites would respond to, and not necessarily a concentration dependency, we tested the essential oils together as a master mix. Additionally, we wanted to test the various individual components by using them from their direct source (e.g., essential oil). Carboxylic acid and aldehyde master mixes were prepared at a concentration of 1 μg/μL. All master mixes were prepared using hexane as solvent. The essential oil master mix contained lavender, eucalyptus, basil, neem, and coriander. However, preliminary data demonstrated that the mites were only responding to some of the active ingredients (linalool, thymol, and eucalyptol), which became the composition of chemicals used for the essential oil master mix at a final concentration of 100 ng/μL. The carboxylic acid series consisted of propionic acid, butyric acid, isovaleric acid, hexanoic acid, and octanoic acid. The aldehyde master mix contained propionaldehyde, pentanal, hexanal, heptanal, octanal, and nonanal.

The GC-FID linked to an GC-ETD was used to record the olfactory response of female T. urticae. The FID reached a temperature of 320°C before ignition. Oven temperature was initially held at 50°C for 5 minutes, then increased to 186°C at a rate of 8°C/min, and then up to 280°C at a rate of 40°C/min where it was held for 2 minutes. The column had a flow rate of 1.2 mL/min, and helium was the carrier gas. Samples were injected in splitless mode. The GC-ETD response was recorded in concert with the GC output using GC-EAD software (GcEad 2012 v1.2.4; Ockenfels Syntech, Buchenbach, Germany) for the duration of the GC run.

After determining which compounds elicited olfactory response in female T. urticae, a novel behavioural assay was developed to determine whether perceived compounds induced attraction or repellency ( Figure 2 ). For each replicate, a leaf from a bean plant was dipped halfway into the compound horizontally (at a concentration of either 10 ng/μL or 100 ng/μL) so that one half from the primary vein on the leaf was treated with compound of interest and the other half treated with the solvent only (i.e., control). Compounds were dissolved in Tween-80 (aqueous) at 0.1% v/v dissolved in water. Concentrations of either 10 ng/μL or 100 ng/μL were tested based on the concentrations to which female T. urticae responded in ETG experiments. A piece of filter paper with a width of 5 mm was adhered along the primary vein of the leaf with a small amount of petroleum jelly (Vaseline^® Petroleum Jelly Unscented, Toronto, Canada). The filter paper strip divided the treatment and control sides of the leaf, and provided a ‘neutral’ launch area for the mite, as well as making it easier to see the mite when transferring to the filter paper. The treated leaf with attached filter paper was then placed between two pieces of plexiglass and held together with a bulldog clip. The top piece of plexiglass had a circular hole cut out with a diameter of 2 cm and a depth of 0.5 cm. A single adult female mite was placed on the filter paper through the hole of the plexiglass and observed for twenty minutes. After twenty minutes the location of the mite was recorded. If the mite was on the control side, the response was recorded as repellent; if the mite was on the treatment side, it was recorded as attraction. In case the mite was still on the filter paper, or escaped the plexiglass, it was recorded as non-responding. Mites had a maximum of twenty minutes to make a choice as it was observed in this study that female T. urticae typically choose a specific location to feed after fifteen minutes.

Prior to testing the compounds of interest, different solvents were tested to determine a viable carrier for all compounds to be dissolved in. A viable solvent would be one to which female T. urticae was neither attracted nor repelled. The potential solvents tested were hexane, ethanol, acetone, and Tween-80 (aqueous) at 0.1% v/v dissolved in water.

Our protocol addressed previous issues with the design of T. urticae behavioural assays, including: (1) we only tested one mite at a time to avoid social behaviour from occurring; (2) we tested potential solvents for repellency prior to testing compounds of interest; (3) we used whole leaves to avoid plant damage-related compounds.

ETG and ETD peak amplitudes were measured on GcEad software (GcEad 2012 v1.2.4 (2012-06-24), Syntech, Kirchzarten, Germany) and exported to Excel (Microsoft Excel for Mac, Version 16.54). To compare sensitivity of female T. urticae to various stimuli, a Shapiro-Wilk normality test was performed on ETD peak amplitudes for each compound at each of the 4 doses (10 ng, 100 ng, 1 μg, and 10 μg). A Two-way Analysis of Variance (ANOVA) with Tukey test was performed to determine if there was a significant difference in peak amplitude between compounds and/or dose (α = 0.05).

To determine potential repellency of compounds through two-choice behavioural assays, binomial tests (performed using the R Stats Package) were performed for each compound at each concentration to determine whether the proportion of mites that choose treatment side was greater than 0.5 (one-tailed test with α = 0.05). All statistical analyses were done using R version 4.0.3 (R Core Team 2020).

Peak ETG amplitudes were significantly different among compounds, as determined from results of two-way ANOVA (F _7,20 = 5.76, p < 0.001) in which peak amplitude of butyric acid was significantly higher than all other compounds ( Figure 3 ). The mean standardized peak amplitude for butyric acid at 10000 ng was 0.923 ± 1.011 mV. Adjusted p-values of butyric acid compared to all other compounds are shown in Table 1 . Peak ETG amplitudes were significantly different between concentrations, as determined from results of two-way ANOVA (F _3,20 = 3.709, p < 0.05). There was a significant interaction between compound and concentration for peak ETG amplitudes, as determined from results of two-way ANOVA (F _21,20 = 3.328, p < 0.001).

Two-spotted spider mites responded to a series of carboxylic acids via stimulation from GC effluent ( Figures 4A , 5A ), including propionic acid, butyric acid, isovaleric acid, hexanoic acid, and octanoic acid. Butyric acid, a 4-carbon carboxylic acid, had a mean peak amplitude of 0.132 mV ± 0.0223. Carboxylic acids with 4 carbons had a smaller mean peak amplitude. Two-spotted spider mites also responded to a series of aldehydes ( Figures 4B , 5B ), including pentanal, hexanal, heptanal, octanal, and nonanal. Heptanal and octanal had a mean peak amplitude of 0.127 mV ± 0.0354 and 0.123 mV ± 0.00901 respectively, and all other aldehydes had a smaller mean peak amplitude. Finally, the mites responded to 3 different essential oil active ingredients (e.g., eucalyptol, linalool, and thymol) ( Figures 4C , 5C ). Eucalyptol had a mean peak amplitude of 0.0973 mV ± 0.0197. Other essential oil active ingredients had a smaller mean peak amplitude. Overall, the mites had the strongest responses to the carboxylic acid series, then the aldehyde series; the weakest response was recorded toward the essential oil active ingredients.

When testing potential solvents for two-choice behavioural assay ( Figure 6 ), female T. urticae was significantly repelled by hexane (p < 0.0001), ethanol (p < 0.0001), and acetone (p < 0.01), as determined from results of exact binomial test. Tetranychus urticae was not significantly repelled by Tween-80 (0.1% v/v), as determined from results of exact binomial test (p = 0.1), so this was determined to be a suitable solvent to use for behavioural assays.

The results of the exact binomial tests determined that T. urticae females were significantly repelled by octanal at 10 ng/μL (p < 0.01), and at 100 ng/μL (p < 0.0001). Tetranychus urticae was repelled by butyric acid at 100 ng/μL (p < 0.001), but not at 10 ng/μL (p = 0.1). Isovaleric acid was repellent at 10 ng/μL (p < 0.0001), but not at 100 ng/μL (p = 0.7). Tetranychus urticae was repelled by heptanal at 100 ng/μL (p < 0.0001), but not at 10 ng/μL (p = 1). Tetranychus urticae was not significantly repelled by linalool at 10 ng/μL (p = 0.4), but was repelled at 100 ng/μL (p < 0.01). Tetranychus urticae was not significantly repelled by thymol at 10 ng/μL (p = 0.6) but was repelled at 100 ng/μL (p < 0.01). Tetranychus urticae was not significantly repelled by eucalyptol at 10 ng/μL or at 100 ng/μL (p = 0.4; p = 0.2). The results of the behavioural assays demonstrate that there was a concentration dependency for the repellency of T. urticae females to various compounds ( Figure 7 ).