Krill clogs the baleen: an unexpected hydraulic constraint on rorqual feeding

How does a fin whale manage to expel sixty cubic metres of water through its baleen in about thirty seconds, when the krill it has just captured lines that filter and should considerably slow the passage of water? A team at Stanford University has measured this resistance in the laboratory. The result, published on 9 July 2026 in the Journal of Experimental Biology, brings out a thirtyfold gap between what is observed at sea and what the physics of flow allows.

The mechanism studied: lunge feeding

Rorquals, among them the fin whale and the blue whale, feed by lunge feeding, known in French as fente d’engouffrement. The animal accelerates towards a swarm of prey, opens its mouth wide and lets the water distend the expandable pouch that runs beneath its jaw and its belly. It then closes the mouth and forces the water out through the baleen, fringed plates of keratin that retain the prey.

This filtration is the rate-limiting step. In a twenty-metre fin whale, one gulp amounts to about 60 cubic metres of water, slightly less than a forty-foot shipping container, and contains between 40 and 144 kilograms of krill, 90 as a median value. The usable baleen area through which this water has to leave is about 2.9 square metres.

The ratio between these two quantities is where the problem lies. Spread over 2.9 square metres, the captured krill forms a layer with a mean thickness of 4 centimetres, ranging from 2.5 to 6.3 centimetres depending on the case. A bed of particles set across a flow behaves as a porous medium: it opposes a hydraulic resistance that grows with its thickness.

Protocol: a theoretical estimate and an experimental measurement

The study combines two complementary approaches.

First strand, the available pressure. The authors estimated the pressure the whale can apply to the contents of its pouch, taking into account muscular contraction and the elastic recoil of the blubber. They obtain 4.1 kilopascals for an average contraction and 17.7 kilopascals at maximum effort. For scale, one kilopascal corresponds to the pressure exerted by a column of water about ten centimetres high: the whale therefore has the equivalent of 40 centimetres of water in ordinary conditions, and 1.8 metres at peak.

Second strand, the resistance of the krill. This was measured directly. Commercially frozen krill was laid out as a bed of controlled thickness in a 1.6 metre tube, in three configurations of 1.9, 3.7 and 5.6 centimetres. Water was pushed through it at pressures of 8.9 to 12.4 kilopascals, and the flow speed measured. The values obtained were then scaled to the dimensions of an adult fin whale.

Results

Observations from on-animal tags establish that a gulp is expelled in about thirty seconds, which implies a speed through the baleen of 0.67 metres per second. The speeds measured through the krill, under average muscular pressure, are of an entirely different order of magnitude.

Thickness of the krill bed	Measured speed	Ratio to the required speed
None (clear baleen)	0.67 m/s	reference
1.9 cm	0.039 m/s	17 times slower
3.7 cm	0.022 m/s	30 times slower
5.6 cm	0.020 m/s	34 times slower

Converted into an emptying time for a full gulp, these speeds give values incompatible with the observed behaviour:

• duration observed at sea: 31 seconds;
• duration calculated at average pressure, 3.7 cm bed: 940 seconds, that is 15.7 minutes, about twice a complete feeding dive;
• duration calculated at maximum pressure, same bed: 132 seconds, still 4.3 times the observed value.

To keep to the 31 seconds measured, a pressure of 25.2 to 30.3 kilopascals would have to be applied depending on the thickness of the bed, above the estimated anatomical maximum. The shortfall is therefore not made up by greater effort: it is structural.

Four explanatory hypotheses

The authors put forward four mechanisms that could resolve the gap, without taking a position on their respective contributions.

1. Resistance in parallel. If about 15% of the baleen surface remains free of krill, the flow concentrates there and the overall resistance drops sharply. The principle is that of two pipes in parallel: the flow preferentially takes the less resistant one. This is the simplest hypothesis, but it presupposes a mechanism that keeps that area clear.

2. Cross-flow filtration. If the water travels parallel to the baleen rather than perpendicular to it, it carries the particles along instead of pressing them against the plates. This principle, used in industrial filtration under the name of tangential filtration, would prevent the bed from forming. What, inside the pouch, would produce such a flow remains to be identified.

3. The action of the tongue. This very bulky organ invaginates into the pouch during engulfment. It could keep the prey in suspension, or dislodge the bed once it has formed. No direct observation documents this role at present.

4. The swimming of the krill. The prey are mobile and could resist being pressed against the plates. Their escape speed, between 0.28 and 1 metre per second, nevertheless generally remains below the 0.67 metres per second of the flow, which limits the reach of this hypothesis without ruling it out.

Limits of the protocol

Three differences separate the experimental set-up from the real situation, and they call for caution.

The krill used was frozen, therefore dead and motionless, which rules out the fourth hypothesis by construction. The bed was of uniform thickness, whereas the actual distribution of prey over the baleen is unknown and probably heterogeneous. Lastly, the tube is rigid and straight, while a rorqual’s pouch is a deformable volume in which the flow is in all likelihood three-dimensional.

These reservations do not call into question the measurement itself, which establishes a physical bound. What they do indicate is that the resolution of the gap probably lies in the aspects the set-up does not reproduce.

Scope of the result

Lunge feeding is one of the most energetically costly actions in the repertoire of the large rorquals: opening the mouth at full speed creates considerable hydrodynamic drag, comparable to deploying a brake. The animal’s energy balance therefore depends on the ratio between the energy contained in the prey captured and the energy spent capturing them, as well as on the number of gulps that can be taken within a single breath-hold. Our page on energy and metabolism in cetaceans sets out this balance in detail.

A filtration time multiplied by thirty would mechanically reduce the number of gulps per dive, and with it the net gain over a day. Identifying the mechanism that prevents clogging therefore amounts to identifying one of the conditions that make gigantism viable in mysticetes. Our pages on diets and interactive anatomy place the filtering apparatus in context.

Reference: Ackermann I. A., Cade D. E., Goldbogen J. A. and Denny M. W., “Constraints on lunge feeding: krill can clog the baleen of filter-feeding whales”, Journal of Experimental Biology, 229(13), 9 July 2026. Press release: The Company of Biologists.