Optimized regularized Stokeslets: calibrating reduced models of multicellular colonies with detailed single-cell models*

Many microbial eukaryotes, like choanoflagellates,  are unicellular, while others form multicellular colonies. However,  the advantages or disadvantages in feeding performance or predator avoidance between single-celled versus multicellular colonies are not yet well understood.  Studying the hydrodynamics of colonies using detailed models of cells that represent flagella, microvilli, and cell bodies would be ideal.  However, given that there are hundreds of cells or more within a colony, this is not feasible.  Reduced models of swimmers are a natural choice.   Here we present a novel approach that uses the flow-field produced by a detailed computational model of a choanoflagellate to find the optimal parameters in a regularized force dipole reduced model that best fits averaged far-field flow.  This optimization also selects the regularized delta function (blob) from a given class.  We find that the optimal blob shape depends upon morphological features of the detailed cell model, like the presence of a microvilli collar. Here we use this approach, along with experimental data, to investigate the hydrodynamic advantages of flagella positioning, cell density, and the influence of colony shape on the swimming and feeding abilities of C. flexa colonies.