The morning a batch is clear is the morning people change four things at once and learn nothing. The useful habit is to read the failure first — the colour, the smell, what is on the bottom, what is on the glass — and then work through the causes in order of how often they actually occur.
Read the failure
| Sign | Most likely cause | Recoverable? |
|---|---|---|
| Pale yellow-green, clearing from the top | Nitrogen limitation — under-dosed or overdue for a split | Usually — dilute 1:1 with fresh medium |
| Clear water, green sediment on the bottom | Culture has died and settled; often heat or chlorine | No — restart |
| Brown or rust-coloured | Senescence or a bacterial bloom | No — discard |
| Green film on the glass above the water line | Biofilm; contamination has established | No — discard and sterilise |
| Cloudy, with a sour or sewage smell | Bacterial contamination | No — discard |
| Clumps or strings in a flagellate culture | Diatom or cyanobacterial contamination | No — discard |
| Dark green, no growth, no clearing | Light-limited, not nutrient-limited | Yes — thin the culture |
The diagnostic order
- Temperature. Not with your hand — with a thermometer, at the vessel, at the time of day the room is hottest. Nannochloropsis tolerates roughly 10–32 °C and dies within hours above 35 °C. A bench by a window can swing 10 °C between morning and afternoon, and a heat spike is the single fastest way to lose a batch.
- Light schedule. Twelve hours on, twelve off. A culture left on continuous light bleaches and stalls; one in a dark corner goes pale and slow. If nobody can say when the light is on, that is the answer.
- Air and gas exchange. A slow stream of small bubbles, a vented cap and headspace. A sealed bottle runs out of carbon dioxide and builds oxygen; a rolling boil strips the CO2 the cells need. Either way the culture stalls, and it looks identical to starvation.
- Water. Distilled or RO, marine salt at 30 ppt, mixed clear before the culture goes in. Tap water is chlorinated specifically to kill microorganisms, and it does not distinguish between the ones you want and the ones you do not. This is the most common cause of a culture that never greens up at all.
- Nutrient. Two millilitres of f/2 per litre, added at every split, not once at the start. Under-dosing shows as yellowing at the meniscus around day four. Slight over-dosing is harmless to the cells, though unused nutrient feeds contaminants.
- Contamination. The one you cannot adjust your way out of. Ciliates, cyanobacterial filaments, a diatom bloom in a flagellate culture, or a film on the glass. It arrives from unwashed vessels, shared pipettes, tap water, or a culture left open to the room. Once it is established, the batch is finished.
Age belongs on that list too, at the end. A culture that has been split weekly for months is not the same culture that arrived: selection favours whatever grows fastest in your conditions, and a line that has been running since spring is more fragile than a fresh tube, even when it looks right.
Do not try to rescue a contaminated culture. You will spend a week, lose the week, and still restart — with the contamination now on your pipette.
Restarting without losing the month
If the culture is dead or contaminated, discard it — down the drain, not into a tank — and sterilise the vessel before anything else touches it. A dilute bleach soak, a thorough rinse with distilled water and air-drying is enough for glass and PET; skip the step and the next batch inherits the problem. Replace the airline if it has been sitting in the failed culture, and do not reuse the cap without washing it.
Then restart from a clean source: a backup vessel, or a fresh 50 mL starter. This is the moment the second bottle pays for itself. A hatchery with a backup splits and carries on; a hatchery without one waits two to four days for a tube, or three to four weeks for a strain grown to order, and either way misses the spawn.
One practical note on timing: restart into 500 mL rather than straight into your full volume. A small restart is cheap to lose and quick to read, and a culture that is going to fail will fail there in three days rather than in your production vessel in seven.
Prevention, which is boring and works
- Two vessels, always. Different ages. A crash in one is an inconvenience, not the end of the line.
- Split weekly, at about half. Do not wait for a culture to look tired before splitting it; by then it is already over-shooting its nutrient.
- Log the Secchi depth. A number written down on day zero, three and five tells you a culture is failing before it is visible. "It looks greener" is not a measurement.
- One pipette per vessel, or none. Cross-contamination between a healthy culture and a suspect one is how a single failure becomes two.
- Clean vessels, distilled water, vented caps. Most contamination enters through one of those three doors.
- A starter on the shelf before the season. The cheapest insurance in this business is 50 mL of a strain you are not currently using.
If you are stuck on a specific failure, send a photograph with the temperature, the Secchi reading and how old the culture is. Questions get answered whether or not an order follows — that has been the rule here from the start.
Restart with a 50 mL starter
Nannochloropsis oculata leaves in two to four days from weekly stock; the other five strains are grown to order. Or order a batch culture if you need the volume now rather than in a week.