What a plankton-net sample can—and cannot—tell you

A net defines the sample before it catches anything

A plankton net is a selective sampler. Its catch depends on the question, the mesh aperture, the net opening and length, and how the gear moves through the water. NOAA’s sampler-design memorandum emphasizes that no single net collects phytoplankton, zooplankton and small nekton equally. A useful interpretation therefore begins with the target organisms and size stages, not with a claim that one tow represents “all plankton.” [1]

Mesh is a sieve, but not a perfect size boundary. Smaller organisms can pass through openings; organisms wider than the openings can still deform, escape or be damaged. A fine mesh may retain smaller forms while clogging sooner, changing how much water passes through it. Net design is a balance among retention, filtration, sample volume and practical tow conditions, rather than a contest to choose the finest mesh available. [1]

Tow geometry gives the catch a depth context

A vertical tow through a planned depth interval combines organisms encountered along that path. An oblique tow crosses depth while the vessel moves, and a horizontal tow samples a more restricted layer. These designs answer different questions. Unless the sampler opens and closes at chosen depths, a catch from a long tow cannot say exactly which part of the path contributed each organism.

NOAA Fisheries’ West Coast methods make the distinction concrete: its documented program uses a 202-micrometer vertical net with a flowmeter from near the seafloor—or a maximum of 100 meters in deeper water—to the surface, and a separate 335-micrometer bongo net towed obliquely over the upper 20 meters for larger groups such as fish larvae and euphausiids. Those are complementary choices for that monitoring program, not a universal recipe for another coast or research objective. [3]

Counts need a sampled-water denominator

A raw count is a count in a jar, not an abundance per cubic meter. For a quantitative estimate, researchers need an estimate of the seawater that actually passed through the net. A calibrated flowmeter can help estimate filtered volume; its reading, calibration factor, net opening and deployment record belong with the sample. NOAA’s field guidance describes combining flowmeter-based volume with organism counts to infer concentration, while also calling for a planned, location-aware sampling strategy. [2]

Consider a deliberately hypothetical tow that catches 120 target organisms and filters an estimated 6 cubic meters. The simple quotient is 20 organisms per cubic meter. That arithmetic is only as sound as the filtered-volume estimate, identification, retention and tow record. If clogging diverts water around the net or slows filtration, nominal distance traveled multiplied by net-mouth area can overstate the sampled volume. Report the method and assumptions so the number remains interpretable. [1]

Mesh and behavior can change the apparent community

A comparative study of several net systems in a Norwegian fjord found that mesh size strongly influenced estimated biomass and species composition. The study reported that about half of organisms with body width equal to the mesh aperture escaped, and that towing speed could add to the loss of smaller organisms. It also found that systems with comparable mesh and filtration conditions could produce similar estimates. These are results from a particular intercomparison, not a correction factor to apply automatically to every species or net. [4]

The sample can also vary because plankton are patchy in space, depth and time. Two nearby tows may encounter different concentrations even if the gear is identical. Larger, mobile organisms may avoid a net, while fragile forms can be damaged; small organisms may be under-retained by coarse mesh. Consequently, a change in observed counts could reflect ecology, collection selectivity, or both. Replicates, method consistency and explicit gear notes help separate those possibilities; none makes the water perfectly uniform. [2]

Compare like with like—and label the limits

For a time series, keep mesh, net type, tow path, depth range, speed, time of day, volume calculation and sample processing as consistent as possible. If a protocol changes, mark the break rather than presenting the resulting counts as a seamless trend. A change from a coarse oblique net to a fine vertical net may reveal more small organisms, but it also changes the question the sample answers. [1] [3]

When comparing studies, check their mesh aperture and weave, mouth diameter, tow depth and direction, flowmeter use, replicates, organism size classes and preservation or identification approach. A number from a net is evidence about organisms retained under those conditions. It is not a census of all life in the water column, a direct measure of ecosystem health, or proof that an undetected species was absent. The strongest reading keeps the sampling window visible beside the biological result. [4] [2]

Sources & further reading

Have a correction or a source to suggest? Send a note →