Abstract
Crustaceans such as crabs, lobsters and crayfish use chemical sensing to determine the location of predators, prey, potential mates and habitat. Chemical signals are primarily distributed throughout terrestrial and aquatic environments by convective and diffusive transport processes that are affected by turbulence. Many animals actively sample odor-bearing fluid using appendages bearing arrays of hair-like chemosensory and mechanosensory sensilla. These animals sample their surrounding odor and fluid environment by flicking their appendages, essentially taking a "sniff". A comprehensive understanding of the chemosensory system in these animals requires examination of the morphology of olfactory appendages, kinematics of flicking behavior and the nature of flow and odorant distribution in the aquatic environments around them. Chemosensing in animals also has the potential to inspire the design of artificial chemical sensors. Numerical modeling and examination of experimental data were combined to examine the nature of turbulent benthic flows and olfactory systems in animals. A numerical model was developed to determine advective-diffusive transport of odorant molecules to olfactory appendages of the crayfish, Procambarus clarkii. I tested the extent of molecule transport to the surfaces of aesthetasc sensilla during an antennule flick and the degree of odorant exchange during subsequent flicks. Odorant molecules were advected between the aesthetascs during the rapid downstroke of the flick and were trapped between the sensilla during the return stroke. Up to 97.6 % of these odorants are replaced with new odorant molecules during subsequent flicks. The concentration of molecules captured along aesthetasc surfaces was found to increase with increased gap spacing between aesthetascs, flick speed, and distance from the proximal end of the aesthetasc. Secondly, to determine how simultaneous flow and odorant sampling can aid in search behavior, a three-dimensional numerical model for the near-bed flow environment was created. A stream of odorant concentration was released into the flow creating a turbulent plume, and both temporally and spatially fluctuating velocity and odorant concentration were quantified. Odorant fluxes measured transverse to the mean flow direction, quantified as the product of the instantaneous fluctuation in concentration and velocity, v'c', show statistically distinct magnitude and directional information on either side of a plume centerline over integration times of <0.5 s. Aquatic animals typically have neural responses to odorant and velocity fields at rates between 50 and 500 ms, suggesting this simultaneous sampling of both flow and concentration in a turbulent plume can aid in source tracking on timescales relevant to aquatic animals. Lastly, the hydrodynamic sensory system and its dependence on sensilla morphology was studied. Numerical models of four predominant sensilla morphologies were simulated to quantify their deformation under the pressure of the incoming flow as a result of upstream flow perturbation. Results show that peaks in velocity and pressure at the sensilla are caused by the upstream perturbation. The standing feathered sensilla are found to experience the maximum force, bending moment and deformation among the tested sensilla morphologies.
