Gas-chromatography-electroantennographic detection (GC-EAD) is a technique used in the identification of volatile organic compounds (VOCs), such as pheromones and plant host odors, which are physiologically relevant to insects. Although pheromones often elicit large EAD responses, other behaviorally relevant odors may elicit responses that are difficult to discern from noise. Lock-in amplification has long been used to reduce noise in a wide range of applications. Its utility when incorporated with GC-EAD was demonstrated previosuly by chopping (or pulsing) effluent-laden air that flowed over an insect antenna. This method had the disadvantage that it stimulated noise-inducing mechanoreceptors and, in some cases, disturbed the electrochemical interfaces in a preparation, limiting its performance. Here, the chopping function necessary for lock-in amplification was implemented directly on the GC effluent using a simple Deans switch. The technique was applied to excised antennae from female Heliothis virescens responding to phenethyl alcohol, a common VOC emitted by plants. Phenethyl alcohol was always visible and quantifiable on the flame ionization detector (FID) chromatogram, allowing the timing and amount of stimulus delivered to the antennal preparation to be measured. In our new chopper EAG configuration, the antennal preparation was shielded from air currents in the room, further reducing noise. A dose-response model in combination with a Markov-chain monte-carlo (MCMC) method for Bayesian inference was used to estimate and compare performance in terms of error rates involved in the detection of insect responses to GC peaks visible on an FID detector. Our experiments showed that the predicted single-trial phenethyl alcohol detection limit on female H. virescens antennae (at a 5.0% expected error rate) was 140,330 pg using traditional EAG recording methods, compared to 2.6–6.3 pg (5th to the 95th percentile) using Deans switch-enabled lock-in amplification, corresponding to a 10.4–12.7 dB increase in signal-to-noise ratio.
All Science Journal Classification (ASJC) codes
- Ecology, Evolution, Behavior and Systematics