Flow disturbances generated by feeding and swimming zooplankton

Abstract

Interactions between planktonic organisms, such as detection of prey, predators, and mates, are often mediated by fluid signals. Consequently, many plankton predators perceive their prey from the fluid disturbances that it generates when it feeds and swims. Zooplankton should therefore seek to minimize the fluid disturbance that they produce. By means of particle image velocimetry, we describe the fluid disturbances produced by feeding and swimming in zooplankton with diverse propulsion mechanisms and ranging from 10-µm flagellates to greater than millimeter-sized copepods. We show that zooplankton, in which feeding and swimming are separate processes, produce flow disturbances during swimming with a much faster spatial attenuation (velocity u varies with distance r as u ∝ r(-3) to r(-4)) than that produced by zooplankton for which feeding and propulsion are the same process (u ∝ r(-1) to r(-2)). As a result, the spatial extension of the fluid disturbance produced by swimmers is an order of magnitude smaller than that produced by feeders at similar Reynolds numbers. The "quiet" propulsion of swimmers is achieved either through swimming erratically by short-lasting power strokes, generating viscous vortex rings, or by "breast-stroke swimming." Both produce rapidly attenuating flows. The more "noisy" swimming of those that are constrained by a need to simultaneously feed is due to constantly beating flagella or appendages that are positioned either anteriorly or posteriorly on the (cell) body. These patterns transcend differences in size and taxonomy and have thus evolved multiple times, suggesting a strong selective pressure to minimize predation risk.

Keywords: biological fluid dynamics; optimization.

Fig. 1.

The study organisms with their diverse propulsion equipment. (A–I) The dinoflagellate Oxyrrhis marina (the other dinoflagellates look similar) (A), the ciliate Mesodinium rubrum (B), nauplius of the copepod Acartia tonsa (the nauplius of Temora longicornis looks very similar) (C), the rotifer Brachionus plicatilis (D), the copepod Oithona davisae (E), the cladoceran Podon intermedius (F), the copepod Metridia longa (G), the copepod T. longicornis (H), and the copepod A. tonsa (I).

Fig. 2.

(A–E) Temporal fluctuations in area of influence, S_{0.1 mm/s}, for the dinoflagellate O. marina (A); peak propulsion speed (B); speed variability index (C); area of influence, S_{0.5 mm/s} during the peak of the power stroke (D); and power of spatial flow attenuation (E), all as a function of Reynolds number for swimmers (red symbols and lines) and feeders (blue symbols and lines). The regression lines in D are as follows: swimmers, Log(S, mm²) = −1.54 + 1.36 Log(Re); feeders, Log(S, mm²) = −0.48 + 1.61 Log(Re). Speed variability index is estimated as the difference between peak and average speed divided by the length of the organism. All data are reported in Table S1.

Fig. 3.

Examples of snapshots of flow fields generated by swimming and feeding zooplankton. (A–F) Swimming Oxyrrhis marina (A), nauplius of Temora longicornis producing feeding current (B), swimming Podon intermedius (C), cruising Metridia longa (D), hovering T. longicornis (E), and repositioning jump of Acartia tonsa (F). The position of the organisms is indicated by red ellipses and the swimming direction by white arrows (gray arrow for the hovering T. longicornis). Flow field animations for all species examined are shown in Movies S2–S4.

Fig. 4.

Examples of the spatial attenuation of flow velocities. (A–H) A. tonsa copepodite repositioning jump (A), O. davisae female repositioning jump (B), P. intermedius swimming (C), A. tonsa nauplii swimming (D), M. longa cruise feeding (E), O. marina cruise feeding (F), T. longicornis nauplius feeding(G), and T. longicornis hovering (H). The solid circles show the attenuation at the peak of the power stroke and the open circles the attenuation during the time leading up to the peak at times given in milliseconds. The solid lines have slopes between −1 and −4 and were adjusted to line up with the far field flow attenuation at the peak of the power stroke. A characteristic far field flow attenuation was somewhat subjectively assigned to each experiment based on how well it compares with the observations; for observations that were between two integer values, we assigned an intermediate value.