Back

How to Become Invisible to Mosquitoes: A Computational Study of Host Signal Collapse

JIANG, Z.-R.

2025-12-11 ecology
10.64898/2025.12.09.693171 bioRxiv
Show abstract

Mosquitoes locate vertebrate hosts by following coherent gradients of carbon dioxide, heat, and humidity. Nevertheless, a persistent empirical observation is that ordinary electric fans drastically reduce mosquito bites, even in the absence of chemical repellents. Here, we introduce a physics-guided, agent-based simulation demonstrating that mosquito host-seeking is not chemically inevitable but physically fragile. The model integrates directional airflow, vortex-induced chaotic mixing, and thermal decoy fields, and tracks 250 autonomous mosquito agents over 600 time-steps under Monte-Carlo sampling with independent randomized initializations. Three aerodynamic mechanisms were quantified. Mild upward airflow already reduces successful host localization by more than 80%, despite not mechanically preventing flight. Increasing vortex circulation produces a continuous, threshold-like collapse, reducing localization probability below 1% at moderate strength. Thermal decoys cause only a linear dilution of success, indicating misinformation alone cannot trigger collapse. A two-dimensional phase map reveals a robust "invisibility region" where airflow and vortex perturbations interact synergistically, eliminating host detection even when neither factor individually reaches threshold. These results show that mosquito host-seeking operates as a gradient-dependent failure system: once scalar fields lose spatial coherence, navigation collapses independent of sensory capability. This provides a quantitative theoretical basis for fan-mediated mosquito protection and suggests that non-chemical, low-energy airflow strategies can induce physical invisibility to hematophagous insects.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.