Thanks again to Felix Hol and @lambrechtslab.bsky.social

Also credits to Zhong Wan, Phd student with Felix Hol at RadboudUMC who carried all the experiments and analysis with Anopheles and was played a key role in developing the method, especially the Dutch version!

More news on that soon!

🥵 This was a lot of examples, yet only scratched the surface of what BuzzWatch can teach us about mosquito behavior. Have a burning question that BuzzWatch could help answer? Start here:

theomaire.github.io/buzzwatch/

and don't hesitate to contact me or Felix!

We tested mosquito responses to multiple daily stimulations over 9 days

Ae. aegypti showed a 2-fold higher response during the day than at night (expected from a day biting mosquito:) ), although spontaneous flight activity without host cues was nearly the same and close to zero at those times !

The movie shows mosquitoes quickly flying to the heated bottom window and attempting to insert their proboscis, demonstrating their heat-seeking response (and that they are not that smart because a heating peltier elt. will provide a good blood-meal..)

As a final application example, we added a simple "host-seeking" module to measure mosquito responses to short pulses of CO₂ and heat at specific times of the day.

As environmental perturbation, we tested the impact of increasing the photoperiod from 12 to 20 hours of light per day.

Mosquitoes quickly adapted with a new complex rhythm but also showed subtle shifts over time.

This is most striking in the PCA graph, where each dot represents a single day!

In these experiments, mosquitoes had no egg-laying site. We thus tested if egg-laying could reverse blood-feeding induced changes.

➡️ Answer: Yes, but only partially.

Overall, the effects of blood-feeding and egg-laying are small, yet consistent and measurable.

We examined the effect of blood feeding 🩸 by monitoring mosquitoes before and after a blood meal.

➡️Following a 2-day period of full rest, we observed subtle changes in their daily rhythm.

➡️Interestingly, these changes were long-lasting and showed no sign of gradual decline.

Fun side quest : in the global pattern method, datasets from various species can be easily compared, here we show for :

➡️Ae. aegypti (Aaa and Aaf)
➡️Ae. albopictus
➡️An. stephensi, gambiae, and coluzzi

PC1 represents day vs. night activity, while PC2 indicates total activity.

To assess this observation, we developed 2 statistical tools

➡️A "local" tool resembling a "barcode" for easy comparison of different features (e.g. sugar feeding)

➡️A "global" tool using dimensional reduction on daily profiles

Both confirmed that "Aaa" is distinct due to increased midday activity!

First, we used the great panel of Ae. aegypti colonies (tested 10!) from @lambrechtslab.bsky.social and found a clear pattern :

➡️the globally invasive subspecies "Aaa" is more active (fly more) during midday compared to the African "Aaf," but not during the morning or evening peaks.

To demonstrate its potentiel, with BuzzWatch multi-scale analysis, we already explored:

➡️Variations in the daily rhythms of Aedes aegypti subspecies. 🌍🌎

➡️Phenotypic plasticity after blood feeding 🩸 or changes in photoperiod. 🔆

➡️Responses to host-cue pulses 🌬️ during day vs. night. 🌓

The BuzzWatch GUI app allows you to quickly reproduce analyses with just a few clicks. It's actually more than a gadget—its speed is great for exploring datasets and conducting "night science" fishing expeditions without predefined hypotheses.

Once all videos are analyzed, we can combine the data to perform multi-scale analysis. This includes examining flight trajectories in seconds, daily rhythms over hours, and long-term trends over days and weeks.

We created a Python GUI app for mosquito video analysis, tracking flying and resting behaviors. By using abundant footage, we easily generate a clear background devoid of mosquitoes and utilize basic blob detection and tracking—no deep learning and manual labelling needed.