Meeting Abstract
During flight, most insects hold their antennae at fixed positions relative to their head. This position depends on information from multiple sensory cues including optic flow and airflow. How does the antennal motor system combine inputs from such disparate sensory modalities to compute antennal position? We addressed this question using computational tools and experiments. From computational models we proposed that the antennal system comprises of two circuits: i) a purely mechanosensory one (Böhm’s bristles mediated) that maintains antennal position (set point) ii) a multisensory one that computes antennal set point from multiple sensory cues. The models also predict the circuits to be autonomous but coordinated. We tested the predictions in the oleander hawkmoth, Daphnis nerii. Moths, bees and other flying insects respond to increasing airflow by moving their antennae forward. We show that the mechanosensory Johnston’s organ (JO) at the base of the antenna senses airflow and sends information to the antennal motor system. In the absence of JO inputs, moths move and maintain position of antennae, but do not respond to increasing airflow by moving antennae forward. Thus, JO feeds into the multisensory circuit that computes antennal position. These data show that computation and maintenance of antennal position are mutually independent, as predicted by our models. Collectively, we show that the antennal system breaks down the task of dynamically responding to multiple modalities into: i) computation ii) maintenance of set point. We argue that this circuitry allows antennae to respond quickly to perturbations while retaining the ability to use slower modalities to determine position. This type of architecture is therefore robust and likely to be the generic architecture for most appendages.
