Using computational fluid dynamics to assess trait functions of life at low Reynolds numbers.

We applied computational fluid dynamic simulations to three-dimensional (3D) computer models of diatoms to assess the effect of trait functions on niche space without the confounding influence of correlated traits. Sinking behavior of phytoplankton was assessed via computer-simulated experiments to test the physics of life at low Reynolds numbers. Specifically, 3D models of Stephanodiscus niagarae were constructed across the middle of the species size range and were placed in simulations to assess variance in the sinking and acceleration rates of the cells. First, we simulated models of anatomically correct cells as a control group. To assess trait function, in this case the function of the spines that encompass the outer rim of each frustule (cell wall), simulations of the model were rerun under the same conditions with the trait removed from the model as the experimental group. We observed that spines served to reduce the influence of outside forces on the cell, specifically the force of gravity, by reducing the sinking and acceleration rates of spined versus spineless models. We also observed that spines increased the range of sinking rates, which increased the dispersal of a population by increasing the range of responses to turbulence. When Hutchinson (1961) presented the paradox of plankton, there was a caveat "it is hard to believe that in turbulent open water many physical opportunities for niche-diversification exist" (p. 141). Herein we have shown that many opportunities for niche diversification are tied to a single trait. By testing trait function in fluid dynamic simulations, we can examine global trends in biodiversity.