Ambient light color does not detectably change the transient, flight-specific DEET response of Aedes aegypti

We asked whether ambient light color changes how Aedes aegypti respond to the repellent DEET. We reanalyzed an unpublished image dataset from the chamber system of Carver (2018): 12 recordings of about 20 mosquitoes each, under six light colors (blue, green, orange, red, white, yellow), with one recording per color in each of two sessions. Each recording had about 10 minutes with bait only and then about 10 minutes with bait plus DEET. We report two robust findings and one negative result. First, the DEET response is short-lived. Pooled across all 12 recordings, flight motion fell by about half in the minute of DEET onset and recovered within about three minutes (ratio of means 1.99, exact sign-flip one-sided p=0.058); over the full halves there was no net difference (ratio of means 1.01, p=0.49). Second, the drop is a loss of flight, not a loss of detection. A separate presence readout, which also counts resting mosquitoes, stayed almost flat (5.6 to 5.1 detections per frame) while flight fell from 0.51 to 0.35 events per frame pair. The chamber is closed, so presence by itself is not a finding; it shows that the animals remained detectable while flight collapsed, and the recovery of flight excludes mortality. Third, we could not show that color changes the response. A detectability control ruled out an imaging artifact. A baseline-activity control then showed that the apparent color ranking follows how much the mosquitoes flew before DEET (Spearman +0.83). Under a null with no color effect, 1.5 of 6 colors pass the cross-session concordance bar by chance. After adjusting for baseline activity and for selection among six colors, no color stood out (p=0.47). Ratio-of-halves suppression measures in chamber assays are confounded with baseline activity, and prospective studies should match activity across conditions or use time-matched within-recording contrasts.