Meeting Abstracts
Dispersal kernel estimation: a comparison of real and virtual particle dispersion in a coastal marine system
Janelle M. Hrycik, Joël Chassé, Christopher T. Taggart, Barry R. Ruddick
Department of Oceanography, Dalhousie University, Halifax, NS, Canada
(Abstract received 03/13/2012 for session A)
ABSTRACT
Early life-stage dispersal influences recruitment and is of significance in explaining the distribution and connectivity of marine species. Motivations for quantifying dispersal range from biodiversity conservation to the design of marine reserves and the mitigation of species invasion. Here we compare estimates of real particle dispersion in a coastal marine environment with estimates provided by hydrodynamic modelling. We do so by using a system of magnetically attractive particles (MAPs) and a magnetic-collector array that provided measures of dispersion based on the time-integration of MAPs dispersing through the array. MAPs were released in a coastal marine location, after which they dispersed through the collector array over a 5 to 7 d period. The same release and environmental conditions were used in a high-resolution 3D hydrodynamic model to estimate the dispersal of virtual particles (VPs). The number of MAPs captured by the collectors and the number of VPs that passed through each corresponding model location were enumerated and compared. We show that VP dispersal reflected several aspects of the observed MAP dispersal, and that comparisons demonstrated model sensitivity to the small-scale diffusivity parameter (Kp). The one-dimensional dispersal kernel for the MAPs had an e-folding scale estimate in the range of 5.19 to 11.44 km, while those from the model simulations were comparable at 1.89 to 6.52 km, and also demonstrated sensitivity to Kp. Variations among comparisons are related to the value of Kp used in modelling and likely related to MAP losses from the water column and (or) shear dispersion acting on the MAPs, a process that is constrained in the modelling. Our demonstration indicates a promising new way of 1) quantitatively and empirically estimating the dispersal kernel in aquatic systems, and 2) quantitatively validating and (or) improving regional hydrodynamic models.