Shields LabTowson University

Shields Lab / Research

Insect taste, smell and repellents

Current work tests plant-derived repellents on spotted-wing drosophila and the house cricket. Earlier work recorded from the taste cells of caterpillars and the olfactory sensilla of moths.

A cluster of ripe and ripening highbush blueberries among green leaves
Vaccinium corymbosum L. Highbush blueberry · Ericales: Ericaceae. Photo: Tom Scavo, CC BY 4.0
A spotted-wing drosophila standing on a ripe red berry
Spotted-wing drosophila: Drosophila suzukii (Matsumura, 1931) Diptera: Drosophilidae. Photo: Timotej Velkov, CC0

Repellents for spotted-wing drosophila

Spotted-wing drosophila, Drosophila suzukii, is an invasive fruit fly. Most fruit flies lay eggs in overripe or rotting fruit, but D. suzukii females have a serrated ovipositor that cuts into ripe and ripening berries, so infested fruit can look ready to pick and be shipped unnoticed. The fly reached the Northeastern United States in 2011, and U.S. losses in small fruit are estimated at $718 million a year (Northeast SARE project LNE22-455R). Growers control it mainly with broad-spectrum insecticides.

Methyl benzoate is a volatile produced by many plants and an FDA-approved food additive, and it repels the fly. In a field trial in highbush blueberry with Rutgers University and the USDA, methyl benzoate dispensers reduced infestation when fly pressure was very high (Gale et al. 2024). The dispensers were the “push” of a push-pull system, which pairs a repellent on the crop with lure traps that pull the pest away; the traps alone did not give comparable protection, and adding them did not improve on the dispensers’ effect.

Methyl benzoate only works at high doses, though, and flies vary in how they respond to it. We are now screening structural analogs that differ from it in the ester group, a ring substituent, or both (Figure 1). Each compound is tested on female flies in a two-choice spatial repellency assay (Figure 2) and ranked by its RD50, the dose that repels half of the flies. Several analogs repelled flies more strongly and more consistently than methyl benzoate at lower doses (Heinbockel and Shields, NCUR 2026). A structure-activity paper is in preparation, and a high-throughput version of the assay that also scores knockdown has been presented at ABRCMS 2025.

Key paper

Gale et al. 2024, Insects 15(1): 47.

Talks on the analog screen are listed under News.

OOCH₃MB
methyl benzoate
OOCH₃NH₂MA
methyl anthranilate
OOCH₂CH₃H₃COE4MOB
ethyl 4-methoxybenzoate
OOCH₃OCH₂CH₃M2EOB
methyl 2-ethoxybenzoate
Figure 1. Methyl benzoate (MB) and three analogs presented at NCUR 2026 as more potent than MB at lower concentrations. Groups that differ from MB are drawn in red.
Solvent Midline Test compound
Figure 2. Two-choice spatial repellency assay, with the solvent control at the left end and the test compound at the right. Twenty-four three-day-old female flies are released evenly on both sides of the midline in an open glass tube; the compound and the solvent are pipetted onto filter paper at the two ends, and the flies on each side are counted at a fixed time. There is a classroom simulation of this assay on the Outreach page.
An adult house cricket standing in dry grass
House cricket: Acheta domesticus (Linnaeus, 1758) Orthoptera: Gryllidae. Photo: Manuel Raab, CC BY-NC 4.0

Essential oils and the house cricket

We screened 27 essential oils against the house cricket, Acheta domesticus. The crickets moved freely in an open arena that held a bottle with the oil treatment inside, and repellency was scored from the share of crickets that entered the bottle, with fewer entries meaning a stronger repellent (Figure 3). Four strong repellents, peppermint, rosemary, cinnamon and lemongrass oils, were then tested at four doses and showed a clear dose-dependent effect, while the synthetic repellents DEET and IR3535 had no significant effect. Our 2026 review in Insects covers essential oils as alternatives to synthetic repellents.

Key papers

Heinbockel, Alzyoud, Raheel and Shields 2026, Insects 17(1): 106.

Heinbockel and Shields 2026, Insects 17(6): 575 (review).

Open arenaBottleOpeningFilter paper with the oilFewer crickets in the bottle = stronger repellent
Figure 3. The bottle-entry assay, schematic and not to scale. Ten crickets move freely in an open arena that holds a bottle with filter paper treated with the oil inside, and repellency is scored from the share of crickets that enter the bottle: the fewer that go in, the stronger the repellent.
A spongy moth caterpillar on a blade of grass, with pairs of blue and red spots along its back
Spongy moth caterpillar: Lymantria dispar (Linnaeus, 1758) Lepidoptera: Erebidae. Photo: Christine Young, CC BY 4.0

Taste in spongy moth caterpillars

The spongy moth, Lymantria dispar (called the gypsy moth in the lab’s earlier papers), feeds on the leaves of hundreds of tree species. Caterpillars have several kinds of taste sensilla on their mouthparts; the main ones for feeding are two on each galea, the medial and lateral styloconic sensilla, each containing four taste receptor cells (Figure 4). We record from them by tip recording, in which a glass electrode filled with the stimulus solution is placed over the sensillum tip; the same electrode delivers the stimulus and records the cells’ action potentials.

These recordings identified cells that respond to sugars, salt, inositol and plant alkaloids (Table 1). One cell in the medial sensillum, the deterrent cell, responds to caffeine, nicotine, strychnine and aristolochic acid, and also to salt (Martin and Shields 2012b); sucrose and inositol, which stimulate feeding, suppress its response (Martin and Shields 2012a). The same cell responds to the insect repellents DEET, IR3535 and picaridin, and each of them deterred feeding. This was the first report of insect repellents detected by a caterpillar’s taste receptor cells (Sanford et al. 2014).

Key papers

Martin and Shields 2012a, Arthropod-Plant Interactions 6: 259–267.

Martin and Shields 2012b, Arthropod-Plant Interactions 6: 519–529.

Sanford et al. 2014, PLoS ONE 9(6): e99924.

Pore at the tip Cone Style; dendrites runthrough it to the pore Four taste receptorcells below the cuticle
Figure 4. Schematic of a styloconic taste sensillum, not to scale. There is a classroom simulation of a tip recording on the Outreach page.
Table 1. Taste receptor cell responses in spongy moth larvae
StimulusResponding cellEffect on feedingSource
SucroseSugar-sensitive cellStimulatesMartin and Shields 2012b
InositolInositol-sensitive cell, medial sensillumStimulatesMartin and Shields 2012b
Caffeine, nicotine, strychnine, aristolochic acidDeterrent cell, medial sensillumDetersShields et al. 2008; Martin and Shields 2012b
Potassium chloride (salt)Two cells in each sensillum, weakly; in the medial sensillum one is the deterrent cellNot testedMartin and Shields 2012b
DEET, IR3535, picaridinDeterrent cell, medial sensillumDetersSanford et al. 2014
A brown marmorated stink bug seen from the side on dark fabric
Brown marmorated stink bug: Halyomorpha halys (Stål, 1855) Hemiptera: Pentatomidae. Photo: Robert Briggs, CC BY-SA 4.0

Brown marmorated stink bug

In 2017 the lab received a $150,000 grant from the USDA for work on the brown marmorated stink bug, Halyomorpha halys, with USDA research chemist Aijun Zhang and research associate Nicholas Larson. The project set out to collect volatiles from the bugs and to test candidate attractants, feeding deterrents and repellents in electrophysiological and behavioral assays, to help design better traps (Towson University News, September 2017).

A tobacco hornworm moth feeding inside a white trumpet-shaped flower at night, the orange spots along its abdomen showing
Tobacco hornworm moth: Manduca sexta (Linnaeus, 1763) Lepidoptera: Sphingidae. Photo: Clifton Ladd, C.W.B., CC BY 4.0

Olfaction in Manduca sexta (earlier work)

Dr. Shields’s work with John G. Hildebrand of the University of Arizona described eight types of sensilla on the antenna of the female tobacco hornworm moth, Manduca sexta. Recordings from single type-A trichoid sensilla, using a panel of 102 volatile compounds and three plant-derived odor mixtures, sorted these sensilla into three functional types.

Key papers

Shields and Hildebrand 1999a, b, Canadian Journal of Zoology 77: 290–313.

Shields and Hildebrand 2001, Journal of Comparative Physiology A 186: 1135–1151.

Shields and Hildebrand 2001, Microscopy Research and Technique 55: 307–329 (review).

Study organisms

Table 2. Insects studied in the lab
SpeciesCommon nameOrder, familyStage studiedWork
Drosophila suzukii (Matsumura)Spotted-wing drosophilaDiptera, DrosophilidaeAdult femalesRepellency of methyl benzoate and analogs
Acheta domesticus (L.)House cricketOrthoptera, GryllidaeAdultsRepellency of essential oils
Lymantria dispar (L.)Spongy mothLepidoptera, ErebidaeLarvae (galea)Taste receptor cell responses to alkaloids, sugars and repellents
Halyomorpha halys (Stål)Brown marmorated stink bugHemiptera, PentatomidaeAdultsVolatiles, attractants and repellents (USDA-funded project)
Manduca sexta (L.)Tobacco hornworm mothLepidoptera, SphingidaeAdult females (antennae)Antennal olfactory sensilla and plant volatiles

Methods

Morphology
Microscopy of antennae and mouthparts to identify, count and describe sensilla.
Electrophysiology
Tip recording from taste sensilla; single-sensillum and antennal recordings for olfaction.
Behavior
Two-choice spatial repellency and knockdown assays with flies, bottle-entry assays with crickets, and feeding assays with caterpillars, each run across a dose series with a solvent control. Fly repellency is summarized as the RD50 and compared with t-tests or one-way ANOVA.