If you want to raise hordes of zebrafish…

…like I do, here’s a useful resource: Regular Care and Maintenance of a Zebrafish (Danio rerio) Laboratory: An Introduction. It’s text and a video guide to familiar procedures.

Unfortunately, it also assumes you have a commercial zebrafish rack, which if you buy the smallest size available, will set you back about $10,000. I’ve just been using an array of 5 gallon tanks, each with its own filter system, which is cheap to set up, but a pain to maintain. I’m going to bite the bullet and build my own DIY flow-through system this summer, though, which I expect will cost a few hundred dollars…which is more in my budget. If you want to follow along, here’s a paper on a Modular, Easy-to-Assemble, Low-Cost Zebrafish Facility — it’ll be something like that. It’s not exactly the kind of show aquarium you’d have in your home, but if all you care about is embryo throughput, it’s loads of fun and lower maintenance than what I’ve been doing so far.


  1. Uncle Glenny says

    When I was a kid, zebrafish were considered one of those difficult-to-breed-in-captivity species…

    n.b. the reading material documenting this was itself relatively old (5-10 or more years), given how rapidly knowledge of these things was progressing.

  2. sawells says

    I’m now visualising a wacky series of misunderstanding and misprints, leading to PZ trying to keep an entire herd of zebras in a series of little tanks.

  3. discus says

    I’d probably rearrange some of the items in the paper, particularly, I’d put the UV filter in line after the pump so *all* the water passes through it before supplying the growth tanks.
    When I did the plumbing for something like this, I’d make liberal use of double union ball valves, which are incredibly useful for maintenance, even if they increase the cost a little.
    It’s also usually thought of as good practice to have the end of supply line pipes emptying back out into the system rather than ending in a “blind alley” (as if some taps are closed off, parts of the pipes can go anoxic, and hydrogen sulphide is not good for growth…).
    The efficiency of biological filtration could probably be increased by creating a second tower filled with bioballs, ideally with air flowing up through it from a fairly large pump or blower.

  4. fredbloggs says

    Rig it up with an aquaponics system and grow your own greens as well!

  5. says

    Yeah, I’ve built something like this before. I plan on relocating the UV filter as you say, and definitely plan to have no dead ends — there’ll be a valve and a return line at the end of each of those output lines, so at the very least I’ll be able to periodically flush everything. Also, I have to be very careful about leaks — would you believe my lab is located above the university bookstore? Whoops. — so everything will be designed so that if there is a major blowout, it’ll all drain into a sink. (oh, yeah, did I mention I’ll have a benchtop system rather than a standalone floor model?)

    I’m also a little concerned that everything in that system is so open — I’ll have a heater in the reservoir tank to bring the water temp up to 28°C, and evaporation will be an issue. I’ve got to at least have a lid on the reservoir to minimize losses.

    Uncle Glenny: You must be thinking of a different fish. Danios have always had a reputation for being easy to breed. When George Streisinger wandered into a pet store in Eugene, Oregon, knowing nothing about fish (he was a phage geneticist), he asked the clerk which fish produced lots of eggs…and the rest is history.

  6. Lars says

    Fiskeeping (or -breeding) FTW.

    Zebrafish may be regarded as difficult to breed in a community tank because the Corydoras and other bottom feeders slurp up all the eggs, and the bethopelagic critters eat what little survives long enough to hatch. A bare-bottom species tank with marbles or some other structure for the eggs to fall into hiding between, is a different matter.

  7. says

    RIght. That’s the one thing you have to watch for in the little cannibals. They’re prolific, spewing out lots of eggs every day, but they’re also caviar fanciers who’ll gobble up their own progeny. We set up filter traps or marbles on the bottom of our tanks to catch the eggs so they don’t get eaten.

    They’re very kinky and put on a live sex show for us in the morning, when the lights come on. The breeding pairs (or trios, or orgies) in the tank will do their water acrobatics, culminating in a cloud of eggs and sperm, and all the other fish in the tank are watching avidly and dive in for a post-coital celebratory snack. It’s all very porny.

  8. says

    Which brings up another concern: last time I had a rack set up, sometimes the unused eggs would escape getting eaten in the marbles at the bottom. Then they’d hatch, and actually manage to live on the detritus in the tank, and some few would grow, swim up, and exit via the coarsely filtered outflow lines to end up in the reservoir tank. Every year I’d have to scoop up these fish of unknown parentage who were living in the big tank at the bottom of the setup. They were entirely accidental and unfortunately undesirable (I had a 29g show tank in the lab, they’d just end up there), but that’s how easy they were to breed — we couldn’t stop them!

  9. blorf says

    So which horde is more important here, the fish or us? I can’t hench properly without a clear hierarchy structure.

  10. marcoli says

    In a former life (grad school) I did research on crayfish regeneration. I helped maintain a crayfish ‘condominium’ with racks of plastic mouse boxes, each fitted with a tubes for an undergravel filter and air, an overflow drain line, and a water supply line that connected to an RO filter system. The appropriate tubes were connected together with lots of T tubes. Every day someone would flush the system with RO water, which would drain out into a sink. It would take occasional tuning to make sure each tank was getting enough air and so on, but it operated 24/7 without serious problems.

  11. Francisco Bacopa says

    You might be able to buy some used Marineland MaRS multi-tank systems. There were a lot of them on the market a few years ago.

    I’ve had more fun breeding T. albonubes than danios. White cloud minnows are about the same size as zebras, but they lay sticky eggs, so more escape predators. They will breed well with the adults around and will even breed in a community tank if there’s a lot of plants and not a heavy filter flow. Breed like crazy in plastic kiddie pools full of hair algae outdoors.

  12. says


    Where do the cephalopods fit into this system?

    Unfortunately nowhere, as zebrafish require freshwater, whereas cephalopods require saltwater. (Cephalopods would be ideal in a whale-shark breeding system though.)

  13. yazikus says

    I’ve really been wanting to build a 20 gallon riffle tank to do a sort of indigenous minnow deal with native plants and such… It will be a little bit on an investment though. It is great how much information is available online- my library has a limited section dealing with fish.

  14. Rumtopf says

    I need me one o’ these here racks for my cherry shrimp o:
    And now I wonder if anyone uses these shrimp in science, as they’re easy to keep, easy to breed(I have so many egg-laden ladies at the moment) and have a fast generation time.

  15. says

    With the water quality here, I struggle keeping freshwater fish happy. Maintaining a salt water aquarium system? Endless sleepless nights, deep anxiety, frequent crises. No thanks.

  16. carlie says

    A utility pump that replaces 75 gallons of water 40 times per hour placed in the stock tank

    That’s…not a typo? The pump is pumping 75 gallons of water every minute and a half?

  17. Lars says

    Sounds like a typo.

    On a different note: The video recommends changing the biological filter media “if it’s very dirty”. I thought that was a no-no. Properly building up a new bacteria flora from scratch takes 4-6 weeks. I’ve always cleaned my biological media in “used” aquarium water to keep as much as possible of the biofilm alive, and never cleaned it too thouroughly, so that the remaining bacteria can repopulate the filter in the shortest time possible.

    But then again I’m a hobbyist, not a scientist.

  18. discus says

    Re: evaporation, you could use a float valve to keep the reservoir automagically topped up. Of course, there’s a cheaper version of a float valve called a “student”… ;) This might, of course, lead to your sopping wet bookshop problem, but with a large enough “oopsie” pan underneath everything, this should preclude any significantly damp bound paper issues. You could also supply the float valve from a reservoir rather than an endless supply of water.
    The university here eventually banned fishtanks in the computer science building after a couple of leaks above the server room…

  19. discus says

    re: evaporation and heat you could also perhaps design a “greenhouse” around the thing with a light wooden frame and appropriate plastic sheeting, but it might look a bit ghetto. But then the best research apparatus always does. Except in physics. Those guys get all the money. Enough to make it look nice and everything.

  20. Lars says

    The container can be covered with a lid made from clear acrylic/hard plastic sheet, 1mm thick or so. Make cutouts for water pipes and such in the corners. As gravity gently shapes the arcylic, condensed water will drip from the middle right back into the tank. That’ll stop/recycle most of the evaporation, and also keep the hardness of the water more stable.

    The only good reasons not to do that, that I can see, is if you want to avoid the acrylic stealing light from the plants, which are non-existent in this setup anyway. Or if you’re just too lazy to put the lid on before leaving the lab.

  21. Sili says

    a few hundred dollars…which is more in my budget.

    Why not Kickstarter the project?

    I don’t know what you’d offer your backers in return, but if those are the costs, you could easily be kept in fish and tanks for decades.