The apparatus measured curiosity, not bites. For each volunteer, or each worn sleeve, the measured quantity was how many mosquitoes entered the device to investigate [8], with the Aedes rig built so the insects could smell an arm they could not reach [6]. That isolates odor cleanly. It also means the endpoint is short-range orientation rather than a blood meal, so the ranking describes who draws attention, not who ends up covered in welts.
The chemistry results rest on a narrower base than the preference ordering. The preference ordering comes from 119 people [3], but the chemistry was read off the small groups at each extreme of attraction for each species [9], which puts the named compounds and bacteria on a narrower base than the ordering itself [3]. Science News describes those groups only as small, without giving their size [9].
There is also an asymmetry in the stimulus. Two of the three species were offered a living arm and the third was offered cloth [1]: Culex quinquefasciatus, which feeds at night, was tested against a nylon sleeve worn for several hours and then presented in the dark [7]. Mosquitoes navigate by visual, thermal and scent cues [15], so an arm and a sleeve are not interchangeable objects. Within a species, the comparison across volunteers holds; across species, the magnitudes should not be set side by side.
The two mechanisms behind these preferences are not the same. Two species leaned toward low total emission [10] and the third toward particular plant-smelling molecules [11], which are different sensory quantities: one is about how much, the other about which. A masking product that lowered a person's overall output could plausibly move them toward the group two of these species prefer. Alicia Showering of BugBiome, who was not part of the work, reads the divergence as separate evolutionary solutions to host-finding, shaped by which hosts a species feeds on and when it hunts [18].
Three bacterial taxa turned up in the high-attraction group for every species [12], a correlation across people rather than proof of cause. Nobody added or subtracted a taxon on a volunteer and re-ran the assay. Kaylee Marrero says the group's next step is working out how skin bacteria make these odors [20], which is the experiment that would license the repellent and trap-lure ambitions DeGennaro sketches [19].
The effect size is not given: Aedes aegypti's preference for male volunteers is reported as slight but meaningful [13], with no effect size attached. And Maria Elena De Obaldia of Regeneron, who has studied human attraction in mosquitoes and was not involved here, notes that human scent runs to more than a thousand chemicals, many of them bacterial products [14], complex enough that whole classes of compounds may have gone unmeasured [16]. Regeneron is a major financial supporter of the Society for Science, which publishes Science News [17].
