Chen Lab Fish Guide

How we keep fish

This is the short version of the standard zebrafish husbandry guide, rewritten in plain language for people working at our racks. Every section links to the full source so you can read more. Where our lab does something specific, it is marked in a yellow box.

The source is The Zebrafish Book by Monte Westerfield, 4th edition, University of Oregon Press, hosted free by ZFIN. Numbers below come from that book unless the yellow box says otherwise. The book is not ours and is not reproduced here. Read it.

1. Water and temperature

Zebrafish do best at 28.5 °C. Outside roughly 25 to 31 °C they are unlikely to breed and embryos may develop abnormally. Check the temperature when anything changes, including covering a tank or moving a heater.

Water quality comes from removing waste. On a standalone tank, siphon debris from the bottom and replace about a third of the water daily. With a filter, replace about half the water at least weekly. Aged tap water works. More consistent water comes from deionized water with a small amount of sea salt added.

Our lab. Our tanks sit on the vivarium recirculating system, so daily water changes are handled by the system. The temperature rule still applies. If a tank is moved off system for any reason, the daily siphon rule applies to it.

Source: Chapter 1.1, Availability and simple care and Chapter 1.3, Water.

2. Feeding

Feed adults at least twice a day. Several light feedings beat one heavy one.
Give enough that every fish gets some and it is all gone within 5 minutes. Food left after that fouls the water.
Vary the food. Dry flake plus ground pellets is the baseline.
For fish you want to breed, live adult brine shrimp is the best food there is.
Do not feed tubifex worms. They can carry disease.

Baby brine shrimp hatch in about 48 hours at 28.5 °C in aerated salt water. Rinse them before feeding. A rough dose is one Pasteur pipette of diluted shrimp per eight adults.

Our lab. Follow the posted feeding schedule for the room. If a tank label says feed extra, that tank gets a third feeding. Write what you did in the registry notes for that line if it was out of the ordinary.

Source: Chapter 1.2, Food.

3. How many fish per tank

With good water flow, the book keeps stocks at about:

1 fish in a quarter gallon tanklet
25 fish in a 5 gallon tank
50 fish in a 10 gallon tank
150 fish in a 30 gallon tank

That works out to about 1.3 fish per liter for adult stock with good flow. More than that and growth slows, water fouls faster, and fish fight. Fewer than that is fine. Growing fish need the lowest densities.

Fish you are conditioning to breed get more room. The book keeps up to 8 females or 16 males per 10 gallons, held in separate tanks and fed 2 to 3 times a day, until the evening before the cross.

Our lab. Our rack has three tank sizes. Small holds 1.4 L, book guideline 1 adult. Medium holds about 2.8 L, guideline 3 adults; this volume is an estimate and still needs measuring. Large holds 9.5 L, guideline 12 adults. When you add a tank, pick the size and the registry fills the volume and shows the guideline. Enter the total and the male and female numbers. A tank over the guideline shows up under Husbandry checks on the home screen.

Source: Chapter 7.2, Densities and Chapter 2.2, Method for maximal embryo production.

4. Setting up a cross

Zebrafish spawn in the morning, shortly after the lights come on. The room runs on a 14 hour light, 10 hour dark cycle, and that light change is the trigger. Set fish up in the late afternoon so they wait until morning to breed.

Which fish to cross

Zebrafish mature at 10 to 12 weeks. Do not cross younger fish.
Fish between 7 and 18 months give the most embryos. Older fish still breed, but expect fewer eggs.
For steady small clutches, use equal numbers of males and females. For the biggest clutches, use about 1 male to 2 females, for example 4 males with 8 females.

Natural cross, the normal method

Put a pair or a small group into a clean breeding container with a few inches of fish water and a layer of clean marbles or a mesh false bottom. The marbles or mesh keep the adults from eating the eggs.
Use roughly equal numbers of males and females.
A few hours after the lights come on, return the adults to their tanks and collect the embryos from under the marbles.
Wash the embryos well after collection. It cuts down on mold.

Expect 50 to 60 percent of setups to produce embryos. An empty cross is normal and not a failure on your part.

Looking after breeders

Do not collect embryos from the same fish more than two days in a row, and take the marbles out for at least a week between collections.
On days a tank is not breeding, feed it protein rich food several times, so the fish rebuild the reserves that make the next clutch.
Move breeding fish into clean tanks every 10 days.

Our lab. We do not swap tanks. We clean the gunk out of the tank every 10 days. Lab tasks on the home screen lists every tank that is due, oldest first. When you finish one, tap Cleaned today on that tank. The registry records who did it and the date, and the tank drops off the list for 10 days.

Why natural crosses and not squeezing

A natural cross is less stressful for the fish than in vitro fertilization and should be the default. Squeezing is for specific experiments only.

What the cross tells you about genotypes

Heterozygote crossed to wild type gives half heterozygotes, half wild type.
Heterozygote crossed to heterozygote gives a quarter homozygous mutants. If the homozygote is lethal, two thirds of the survivors are carriers.

Our lab. Every line page shows two flags before you cross. Rest, which says whether the fish have had their week. And Age for breeding, which says too young, mature, prime, or past prime, from the book's windows above. The Ready to cross list on the home screen only shows lines that pass both.

When you set a cross, log it in the registry the same day. Enter the date and how many males and females you set. When you collect, add the egg count. The next day, add how many are alive past 1 dpf. The registry works out survival for each cross and for the line, and the Breeding report on the home screen adds it all up for the lab. A line whose last three crosses average under 50 percent survival or under 30 eggs gets a fertility check flag, so you look at those fish before you cross them again.

Source: Chapter 2.1, Steady low level production, Chapter 2.2, Maximal embryo production, Chapter 2.5, Breeding schedule, Chapter 2.6, Breeding over marbles, and Chapter 2.7, Pair wise breeding. The 50 percent and 30 egg fertility thresholds are this lab's own starting point, not the book's.

5. Resting fish between crosses

Handling fish stresses them. The book calls for about a week of rest between natural crosses with the same fish. If you do not need embryos for a while, take the marbles out and feed less, or the fish get fat and breed poorly.

Our lab. Minimum 7 days between crosses, and 8 to 9 days is the sweet spot. The registry shows this on every line: REST under 7 days, OPTIMAL at 7 to 9, READY after that. Past about two months without a cross, fish start losing fertility and may need one or two warm up crosses before they lay well. The registry marks those lines STALE.

Source: Chapter 2.7, Pair wise breeding, Chapter 2.2, and Chapter 2.1. The book gives only the one week minimum. The 8 to 9 day sweet spot and the two month fertility note are this lab's own practice, not the book's. For comparison, UCSF's IACUC procedure calls one week the minimum and two weeks ideal.

6. Keeping a mutant line alive

Keep working stocks as intercrosses of identified heterozygous parents. That gives the most identified carriers per clutch.
Also outcross heterozygotes to the wild type main line regularly, so the background does not drift.
Pool wild type progeny from several crosses, not one, to keep genetic diversity.
Make a new generation every 4 to 6 months. Use one generation for experiments while the next is growing up.

Our lab. Genotype every new generation before it goes on the system as a known line. A tank marked NOT GENOTYPED in the registry is not a working stock yet. Lines borrowed from the core are marked ON LOAN and are never discontinued without checking with Aren or Dr. Chen first.

Source: Chapter 7.2, Guidelines for maintaining mutant stocks.

7. Raising embryos to adults

Day 0 to 3. Keep 15 to 25 embryos per 100 mL of system water, at least 3 cm deep, at 28.5 °C. They hatch at 2 to 3 days.
Day 4. Start feeding live paramecia or a similar infusion. The yolk is nearly gone by day 6, so feed generously from then.
Day 9. Move to 1 to 2 liter containers. Add baby brine shrimp and phase the paramecia out over the next days. Clean twice daily with fresh system water.
Day 21. Move to tanks on the system. Keep feeding brine shrimp and start adult food as soon as they can eat it.

For the fastest growth keep numbers per container low, feed twice daily, and change about a third of the water each day.

Our lab. Embryos stay with us at least 5 days before anything goes on the system. When a clutch goes on, it gets the next free LA number and a registry entry that day. The registry will not accept a line placed on the system fewer than 5 days after its date of birth.

Source: Chapter 3.1, A simple method for raising babies and Chapter 3.2, Raising larvae in a nursery. The 5 day minimum is this lab's rule.

8. Telling males from females

Males are longer and slimmer, with a yellow tint, strongest on the belly. Females are plumper and more silver. Look at the belly, not the fins.

Our lab. Split tanks are labeled A and B. The sex of each tank is recorded in the registry from the counts you enter, so count carefully.

Source: Chapter 2.1.

9. What goes in the registry

The book's advice on record keeping is simple: write down what was crossed, when, what came of it, and who did it. Our registry is that notebook, shared and searchable.

New clutch on the system: add the line with its LA number, parents, and date of birth.
Set a cross: log it that day with the ratio of males to females.
Collect eggs: update the cross with eggs laid and how many made it past 1 dpf.
Screen or genotype: tick the box on the line.
Count a tank: enter total, males, females.
Anything unusual: write it in the notes on that line.

Every change records who made it. Nothing is deleted, only marked discontinued.

Source: Chapter 1.8, Keeping track of stocks.

Full book

Everything above is a summary. The full text, including disease identification, quarantine, cleaning tanks, and recipes for fish water and tricaine, is at zfin.org, The Zebrafish Book. Read the disease chapter before you need it.

Citation: Westerfield, M. (2000). The zebrafish book. A guide for the laboratory use of zebrafish (Danio rerio). 4th ed., Univ. of Oregon Press, Eugene.