Reef Cloverson / Coral science

Does still water change what Acropora releases?

September 24, 2026 · Reef Cloverson

A 2024 flume experiment found that stagnant water produced much larger transparent-particle release and bacterial growth than flowing conditions, but only two coral sets were tested.

Does still water change what a branching coral releases into the surrounding seawater? Wu and colleagues (2024) placed groups of Acropora colonies in a recirculating flume and compared four steady flow speeds. The largest biological shift appeared at the boundary between moving and motionless water, not among the three flowing conditions.

Study at a glance: two groups containing 14 and 20 colonies were each exposed to 0, 5, 10 and 30 cm/s for 24 hours. The work was conducted at the Okinawa Institute of Science and Technology in Japan.

No flow produced the sharpest change.

Transparent exopolymer particles, or TEP, are sticky carbohydrate-rich particles found in coral mucus and seawater. The team tracked the change in TEP concentration, bacterial abundance and the hydrodynamics above a canopy of live colonies.

01Four flow levels
0, 5, 10 and 30 cm/s
EXPERIMENTAL CONTRASTEach coral set passed through the three flowing treatments in randomized order. The no-flow exposure was deliberately reserved for last because the researchers expected it could damage the colonies.
026.2–9.3× higher
Net TEP release at no flow
PARTICLE RESPONSEFor both coral sets, the estimated net release under stagnant water exceeded the average response across 5–30 cm/s. Differences among the three nonzero speeds were small.
033.6–5.1× higher
Bacterial growth at no flow
MICROBIAL RESPONSEBacterial abundance rose in every 24-hour exposure, but the increase was greatest without water movement. Across the pooled observations, TEP release and bacterial growth were positively associated (Kendall τ = 0.71, P = 0.014).
04Visible tissue stress
Observed only during the stagnant trials
QUALITATIVE OBSERVATIONSome branches whitened, lacked visible polyps and shed tissue. This was not a randomized survival endpoint, and the experiment cannot separate a no-flow effect from every possible order or carryover effect.

Three bottles do not equal three independent flow experiments. The triplicate water samples characterized one shared flume exposure. The biologically meaningful replication was two repeatedly used coral sets, both tested in the same apparatus.

WHAT WAS MEASURED
TEP concentration, bacterial abundance and particle-image-based flow fields before or after 24-hour flume trials.
WHAT WAS NOT ESTABLISHED
A universal minimum flow, long-term coral growth, the molecular trigger for mucus release, or performance under wave-driven and changing flow.

What remains unresolved?

Strengths: The study paired live, colony-scale canopies with direct flow measurements. No-coral trials were run for every speed to estimate background changes in the seawater, and the overall pattern appeared in two different canopy arrangements.

Limits: Only two coral sets were tested, each set contained colonies that shared the same water, and the same flume served every condition. The no-flow trial was always last. The design therefore supports a strong within-system observation but provides limited independent replication for generalizing the size of the effect.

The experiment used one day of steady, one-directional flow in a closed 406-L circuit. Natural reefs and aquaria often produce oscillating, pulsed and spatially uneven flow. The authors measured particle and bacterial changes rather than coral energy balance, calcification or recovery.

What this means

Authors’ interpretation: The canopy geometry promoted exchange once water was moving, and stagnation coincided with much larger TEP release and bacterial growth. They proposed inefficient exchange and coral stress as plausible explanations for the stagnant-water response.

Reef Cloverson interpretation: Flow should be treated as a measured environmental variable, not merely a pump setting. This experiment also suggests that “some flow” and “more flow” may pose different questions: the large transition here was from zero to nonzero movement, while 5–30 cm/s produced similar particle and bacterial responses over 24 hours.

That pattern does not prove that 5 cm/s is sufficient for every coral, colony shape or aquarium. A useful next experiment would independently replicate tanks and compare a defined low-flow treatment with a realistic variable-flow treatment over longer periods.

Could we replicate this?

Published protocol: The manipulated factor was steady flow speed at 0, 5, 10 or 30 cm/s. A no-coral run at each speed served as the background control. The same coral set was reused across speeds, with at least one week between trials; two coral sets provided the study’s main biological repeat. Each trial lasted 24 hours at 25°C in filtered natural seawater under a 12-hour light cycle near 200 µmol photons m⁻² s⁻¹.

The full protocol requires a 406-L recirculating flume, controlled impeller, stable lighting and temperature, an Alcian-blue colorimetric TEP assay with absorbance read at 787 nm, DAPI-stained bacterial counts by epifluorescence microscopy, and particle image velocimetry. Its PIV system used a laser sheet, seeded particles and high-speed imaging; that hydrodynamic component is not a routine aquarium measurement.

Proposed smaller pilot—not performed in the paper: Ask whether short, quantified periods of very low flow increase a mucus-particle proxy and bacterial abundance relative to a matched flowing condition. The hypothesis would be that reduced exchange produces a larger response. Use separately maintained tanks or flumes as the independent units, assign flow treatments randomly, keep coral material balanced across tanks, and define stop criteria before exposing corals to near-stagnant water.

A pilot without the chemical assay could document flow, polyp behavior and visible tissue condition, but it could not reproduce the paper’s TEP result. A pilot in ordinary aquaria also would not reproduce a uniform flume velocity. Species identity, colony geometry, flow direction and prior exposure would remain important constraints.

Methods: How they did it

Okinawa, Japan · branching Acropora · 14-colony and 20-colony sets · four 24-hour exposures per set.

01
Build the canopy

Arrange one coral set across a 98 × 30 cm test section in roughly 30 cm of water and preserve that layout across flow treatments.

02
Run the exposure

Replace the seawater for each trial, apply one steady speed for 24 hours and run a matched no-coral control within the next day.

03
Measure change

Quantify TEP and bacterial abundance from start and end samples, then map velocity and turbulent shear separately with particle image velocimetry.

Funding source

The paper reports support from the Okinawa Institute of Science and Technology Marine Biophysics Unit; JSPS KAKENHI grants JP20K19960, JP23K17033 and JP23KJ2133; and JST FOREST Program grant JPMJFR2070. The acknowledgments separately credit OIST imaging and marine-science staff and other technical contributors.

Sources

Wu et al. (2024) · Frontiers in Marine Science 11:1404526 · Primary paper

Japan Society for the Promotion of Science research record · Grant and publication cross-check

Study graphic
Research summary: 2 Acropora sets were tested at 0, 5, 10 and 30 cm/s for 24 hours. No flow produced much higher TEP release and bacterial growth; one flume and two reused coral sets limit inference. View full-size graphic

The research continues here

Open Discussion

Questions, corrections, or a different interpretation of the study? Join the discussion.

Discussion guidelines

Keep discussion respectful and science-based. Focus on the evidence, methods, and interpretation—not the researcher’s character. Support factual claims with sources when possible. No spam, personal attacks, or copied passages from publications.

0 approved comments

No comments yet. Start the discussion.

Leave a comment

Choose the name you want readers to see. Do not use your email address.

Your email will not be displayed publicly. No account is required.

Up to 5,000 characters. Fields marked * are required.

All comments are reviewed before they appear publicly.