Low light versus CO2 in a nano tank
One of these grows plants faster. The other one is much harder to get wrong in eight gallons of water. That trade is the whole decision.
We earn a commission if you buy through the links on this page, at no extra cost to you. It never changes a ranking, and there are products recommended here with no affiliate link behind them at all. Full disclosure.
The short answer
Run low-tech. Every plant this site recommends is published by its seller as needing no CO2, and the whole point of a nano tank is that it is a small, stable, low-effort system. Injected CO2 turns it into a small, fast, twitchy one.
That is a recommendation, not a rule. There is a real case for CO2 in a nano, and it is at the end of this page. But the default answer for the tank on your desk is the boring one.
What CO2 actually does
Plants build tissue out of carbon, and in most aquariums carbon is the limiting nutrient rather than light or fertilizer. Inject CO2 and you remove that limit: growth accelerates, stems get denser, some species color up, and carpeting plants that creep along in a low-tech tank actually carpet.
The catch is that plants only use the extra carbon if light and nutrients keep up. That is the well-known triangle - light, CO2, nutrients - and in a high-tech tank you are balancing all three continuously. Get it wrong in one direction and the plants stall; get it wrong in the other and algae takes the surplus.
The nano-specific problem: no buffer
Everything that makes a small tank convenient makes CO2 harder. A rated 10-gallon holds 8.5 gallons of actual water; a rated 5-gallon holds 4.3. Injecting a gas that lowers pH into that volume moves the number quickly, and a solenoid that sticks open at 2am has very little water to dilute it into.
This is the same argument this site makes about heater wattage: the failure mode that matters in a nano is not the equipment underperforming, it is the equipment failing on, with no thermal or chemical mass to absorb the mistake.
| Low-tech | Injected CO2 | |
|---|---|---|
| Growth rate | Slow to moderate | Fast |
| Plants available | Rhizome, crypts, moss, dwarf sag | Effectively all of them |
| Trimming | Every few weeks at most | Weekly, forever |
| Failure mode | Plants sulk | pH swing, gassed livestock |
| Equipment in the tank | Light | Light, diffuser, drop checker, tubing |
| Ongoing cost | Fertilizer, occasionally | Gas refills, fertilizer, more of both |
| Vacation-proof | Yes | Not really |
What low-tech actually means
Not “no light.” Low-tech means matching the light to plants that can live at that level, and then leaving the system alone long enough to stabilize.
Aquarium Co-Op publishes anubias, java fern, java moss and Cryptocoryne wendtii at 10-20 PAR with no CO2 requirement, and dwarf sagittaria at 20-35 PAR, also with no CO2. That is the entire plant palette this site recommends, and every one of those numbers is a published figure rather than our estimate - the ranked list is here.
The one number nobody publishes is how much PAR your fixture delivers at your substrate. No nano light manufacturer publishes PAR at any depth. We record that absence on every light we list, and it means low-tech planning is done by plant response rather than by measurement - what a nano fixture realistically delivers works through what you can infer instead.
The lever most people should pull instead of CO2
Less light, not more carbon. The commonest low-tech failure is a tank running a bright fixture for ten hours a day with slow-growing plants that cannot use it, which is a description of an algae farm.
- Shorten the photoperiod before you brighten the light. Six to eight hours is plenty for the plants on our list.
- Add plants rather than nutrients. More plant mass competing for the same nitrate is the cheapest algae control there is.
- Feed the fish less. Uneaten food is the nutrient input you control most directly, and in a nano it dominates.
- Root tabs for the root feeders. Crypts and dwarf sag feed through their roots; a liquid fertilizer does little for either in inert gravel.
Liquid carbon is not CO2
Bottled “liquid carbon” products are not dissolved CO2 and do not do the same job. Their manufacturers are generally clear that they are algaecides with a secondary carbon claim, and some plants - certain mosses and vallisneria in particular - are widely reported to react badly. We are not going to tell you they do nothing, because we have not tested them. We will say that if you are buying one to substitute for injected CO2, the honest comparison on this page is between low-tech and real CO2, and liquid carbon sits in neither column.
When CO2 is genuinely the right call
Three cases, and they are narrower than the hobby suggests.
You want a real carpet. Dwarf hairgrass, Monte Carlo and the rest are the plants people picture when they say carpet, and in a low-tech nano they mostly fail slowly. Dwarf sagittaria is our low-tech substitute and it is a grass, not a lawn.
You want red and demanding stem plants. Color in most red species is a function of light and carbon availability. Low-tech gets you green.
The tank is a scape rather than a home. A display tank with no fish, or with hardy shrimp, run by somebody who enjoys weekly trimming, is a completely reasonable place for CO2. The risk profile changes entirely when there is nothing in the tank that can be gassed.
If you do run CO2 in a nano
We are not going to write a setup guide for equipment we have not used, and this site does not list CO2 gear for that reason. What we will say is that the two decisions that matter are both about failure: run it on a timer that shuts off before the lights do, and use a drop checker so the tank tells you what the dissolved level actually is rather than trusting a bubble count.
And do it in a tank that is already established. Adding CO2 to a new nano is adding a fast variable to a system that has not found its slow ones yet.
What we cannot tell you
What bubble rate suits your tank, how any particular regulator behaves, or whether a given liquid carbon product harms a given moss. Those are questions that need equipment we do not own and testing we have not done. Our method is to publish what manufacturers and growers publish, note what they do not, and be plain about which parts of a subject we are not in a position to answer.
What we would buy for this
Chosen from the same registry as our roundups, on published specifications and the arithmetic they imply - not on testing we have not done.
Fluval C.O.B. Nano LED, 6.5W
Six and a half watts, no app, no schedule, no color channels. For a low-light betta tank growing anubias and java fern that is not a compromise, it is the correct specification - and it costs a fraction of the programmable option. Put it on a mechanical outlet timer and you have a photoperiod.
Skip it if: You are growing carpeting plants or anything Aquarium Co-Op rates above low light. Six and a half watts will not do it.
Fluval Plant 3.0 Nano Bluetooth LED, 15W
Fifteen watts of programmable full-spectrum light over a nano tank, with a 24-hour ramp you set in an app and six independently controlled band waves. It is more light than a low-tech betta tank needs and exactly the right light for a planted one. The IP67 rating is the underrated part: it can sit directly over open water without a lid.
Skip it if: You will not use a phone to run a light. There are no physical controls at all - Fluval states it is operated exclusively via the app.
Anubias Nana Petite
The single most useful plant in a betta tank. It attaches to hardscape rather than rooting, it wants low light and no CO2, and its leaves are stiff enough to hold a sleeping betta. Aquarium Co-Op publishes the numbers that matter: 10-20 PAR, no CO2 requirement, attach to rocks or wood.
Skip it if: You intend to bury the rhizome. Anubias rots when the rhizome is planted, and this is the most common way people kill it.
Cryptocoryne Wendtii
The rooted midground plant that actually belongs in a small tank, and the one that will scare you first. Aquarium Co-Op publishes low light at 10-20 PAR, no CO2, an easy skill rating and a mature height of around 8 inches - which is most of the water column in a 10-gallon and all of it in a 5. It also melts back to nothing within a fortnight of being moved, which is normal, is called crypt melt, and is not you killing it.
Skip it if: The tank has bare glass or an inch of decorative gravel and you do not intend to add root tabs. This plant feeds from its roots.
Frequently asked
Do I need CO2 for a nano planted tank?
No. Every plant we recommend is published by its seller as requiring no CO2, including the carpeting one. CO2 makes plants grow faster and gives you access to demanding species; it is not what keeps low-light plants alive.
Is CO2 dangerous for a betta?
It can be, and the risk is higher in a small tank than a large one because there is so little water to dilute an overdose. If you run CO2 in a tank with a betta in it, run it on a timer that stops before the lights go out, and use a drop checker rather than counting bubbles.
Why are my low-light plants growing algae?
Usually too much light for too long against plants that cannot use it, plus nutrients from overfeeding. Shorten the photoperiod to six to eight hours, add more plant mass, and feed less. Reach for a chemical only after those three have had a month.
Does liquid carbon replace CO2?
It is a different product with a different mechanism, and the manufacturers themselves generally position it as an algae treatment. We have not tested any of them, so we will not tell you it works or that it does not - only that the real comparison on this page is low-tech versus injected CO2.
What light do I need for a low-tech nano?
One that keeps the plants on our list alive at 10-20 PAR, run for six to eight hours a day. No nano fixture publishes PAR at depth, so what you are really buying is a light with a timer and a dimmer so you can adjust by plant response.