Fundamentally, if a microorganism is swept away from the vents into the deep sea, it is deprived of highly-concentrated, vent-specific, dissolved food sources; its rate of gain of energy and mass -- and consequently its fitness -- quickly drops to zero. I would like to examine this fundamental and other consequences of fluid motion in three hydrodynamic regimes likely available to vent system microbes: high velocity vents, low velocity tube worm beds, and near stagnant pore spaces in basalt.
First, consider a chemoautotroph attempting to acquire solutes from the flow within a sulfide rock, a pipe-like structure culminating in a (atypically low-temperature, for our purposes) "black smoker" plume rising into the stratified deep ocean water. There, Reynolds number for a characteristic velocity and geometry is about 10,000; the flow is highly turbulent. If the resident microbe has means of attaching itself to the rock, it will enjoy high gradients of nutritious and energetic solutes, constantly refreshed at its cell surface and independent of diffusive processes. Such a strategy is complicated, however, by the risk of either being detached or suffering other highly variable conditions (temperature, for example) which exceed tolerance ranges.
A less turbulent environment is the tube worm bed, at the base of which Reynolds number will be about 10. In this situation, the risk of advection into the deep sea is substantial, so attachment is still crucial, but the less turbulent flow may result in boundary layers that influence the fitness of chemoautotrophs by changing the solute concentration gradients.
A final environment available to subsurface microorganisms is the porous media of basalt itself. Flows within the rock probably exist with varied geometries, but reasonable estimates suggest that Reynolds number may be very low, around 0.01; viscous forces will dominate motion, and diffusive processes will become limiting. Any flow of fluid will be laminar, boundary layers will be thick, and the concentration gradients with respect to the distance from a microbe should be governed by Fick's first law Jumars, 1993).
Because thermal convection in a porous medium results in re-circulatory fluid flow, the residence time of chemoautotrophs in the subsurface may be significantly longer than their reproductive lifetime. Indeed, the center of a convective cell is an advective safe haven! With the diminished risk of being swept out of the system, or at least with the likelihood of having enough time to reproduce before being expelled, and at these very low Reynolds numbers, microbes may benefit from swimming strategies which optimize their position in the subsurface solute concentration gradients (Purcell, 1977). The most fit autotrophs will optimally balance the cost of detaching (being flushed from the system), with the benefit of being free to follow concentration gradients through simple swimming algorithms (if it's getting better, don't stop swimming so soon).
The nature of fluid flow in porous basalt (that likely underlies tube worm beds) results in an intriguing suite of microbial niches, enabling higher potential fitness than in more energetic flows (like those culminating in high-velocity vents). Do vent microbes display flagella and other swimming morphologies? Are their sizes and swimming velocities great enough to elude the limits of local diffusion? Are they chemotaxic, moving like motile E. coli? The combination of a survey of vent chemoautotroph morphology with these fluid dynamic considerations may help to explain the observation of higher abundance and diversity of hyperthermophilic microbes from diffuse flow fluid samples than from focused flow fluid samples (Baross, 1997).
Jumars, P.A. 1993. Concepts in Biological Oceanography: An Interdisciplinary Primer. Oxford University Press, pp. 57-60.
Purcell, E.M. 1976. Life at low Reynolds number. Am. J. Phys., Vol. 45, No. 1, January.