Do Planted Tanks Need CO2? What Actually Changes With and Without
Research-based answer: what CO2 changes (growth rate, species range, carpets), the three paths — low-tech, liquid, injected — and when gas is wasted money.
Table of Contents
Affiliate Disclosure: We may earn a commission if you purchase through our links. This does not change the price you pay. How the desk is funded: How We Make Money.
This is a research-based guide, not a hands-on test. The plant lists, hour bands, and cost logic below come from published planted-tank horticulture sources — the 2Hr Aquarist's CO2 methodology, Aquarium Co-Op's keeper guidance, Diana Walstad's ecology-first writing, and public plant profiles from the major aquatic-plant retailers — cross-checked against the photosynthesis biology both camps agree on. No cylinder was emptied for this page. When two credible sources disagree, both positions get stated.
The question is not "does CO2 help plants." It does; that is not in dispute anywhere in the literature. The question is whether your tank is short of carbon, and that depends on two things you already own or have already chosen: your light and your plant list. Answer those first and the CO2 decision answers itself.
The short answer, by setup
| Your setup | Do you need CO2? | Why |
|---|---|---|
| Easy plants (anubias, java fern, crypts, vals), entry LED, 6–8 hour photoperiod | No | Carbon in tap water plus fish respiration covers these species; gas buys you little but algae risk |
| Same easy plants, high-output light (Chihiros WRGB II class) wide open | Effectively yes | Strong light raises the carbon demand; without gas the imbalance feeds algae, not growth |
| Carpet plants (dwarf baby tears, dwarf hairgrass as a lawn), any light | Yes, injected | The dense-growth species that make actual carpets are the classic high-demand case |
| Medium stems (rotala, ludwigia) you want dense and colored | Nice to have | They survive low-tech but stay leggy and green; injection tightens growth and brings out reds |
| Red plants as the main show (alternanthera, rotundifolia 'H'ra) | Recommended | Most true reds need strong light, and strong light without CO2 is the algae kit |
| New tank, first month, any plants | No | Everything is a new variable already; the photoperiod guide comes first |
Read the table honestly and one pattern shows: CO2 is not a plant question or a tank question. It is a light question. Gas lets plants use photons they otherwise cannot; if your light is modest, the gas has nothing to do.
What CO2 actually does (one paragraph of biology)
Plants build tissue from carbon. In water, the available forms are dissolved CO2 and bicarbonate, and the ambient level in a non-injected tank sits far below what the same plants would pull from air. Aquatic plants still grow — they are adapted to scarcity — but photosynthesis runs carbon-limited, which caps growth rate no matter how much light or fertilizer you add past that point. Injection raises dissolved CO2 toward the level where light and nutrients, not carbon, become the limiting factors. That is the whole mechanism: CO2 does not feed plants directly so much as unblock the growth your light is already offering. It also shifts the competitive balance. Fast-growing plants outcompete algae for resources when carbon is plentiful; carbon-starved plants cannot, and algae, which is never carbon-starved, takes the surplus. This is the same light–carbon–nutrient triangle the light schedule guide works from, viewed from the carbon corner.
The three paths, and what each costs per year
Low-tech, liquid carbon, and pressurized injection are not three quality tiers of the same thing. They are three different commitments.
| Path | Upfront (hedged) | Ongoing (hedged) | Growth ceiling | Failure mode |
|---|---|---|---|---|
| Low-tech / Walstad | Soil-based bed, modest light; often nothing beyond what you own | Occasional ferts; water changes | Slow, steady, easy-species range | Impatience; picking high-demand plants anyway |
| Liquid carbon (glutaraldehyde "Excel" products) | A bottle and a routine | A bottle every one to two months, roughly | Slightly above low-tech; suppresses some algae | Dosing drift; not a real carbon source at scale; sensitive plants and some inverts react badly |
| Pressurized injection | Regulator + solenoid kit, cylinder, diffuser — commonly into the low hundreds once | Refills; a 2.5–5 lb bottle lasts a small tank many months | Highest; unlocks carpets, dense stems, reds | Imbalance (too much light or too few nutrients with the gas on); equipment leaks |
The honest annual-cost comparison: liquid carbon looks cheap for one small tank and gets expensive per gram of carbon compared to a refillable cylinder. Pressurized costs the most on day one and the least per unit of actual carbon, which is why almost everyone who stays in the hobby past the carpet-plant stage ends up there. The hardware specifics of that path — dual-stage regulators, solenoids, thread families — live in our best aquarium CO2 kit guide.
The Walstad method deserves its own sentence because it is the strongest no-CO2 answer: soil-based substrate, moderate light, heavy planting from day one, and the tank's own biology (fish respiration, bacterial breakdown in the soil) supplying carbon. Walstad's writing is the reference for it, and it works — within its plant list. It is an ecology you join, not a downgrade.
Plant picker: what thrives without injection
The single best way to avoid needing CO2 is to pick plants that never wanted it. The species below are repeatedly listed as low-demand across public plant profiles and both major methodology camps:
| Thrives low-tech | Tolerates low-tech but grows slowly | Wants injection |
|---|---|---|
| Anubias (all forms) | Vallisneria | Dwarf baby tears (HC Cuba) |
| Java fern and relatives | Cryptocoryne (most; slow by nature, not by suffering) | Glossostigma |
| Java moss, Christmas moss | Amazon sword (heavy root feeder; wants rich substrate) | Dwarf hairgrass as a true lawn |
| Bolbitis (African water fern) | Rotala rotundifolia (green, leggy form) | Alternanthera reineckii |
| Water sprite, hornwort (floaters generally) | Ludwigia repens (green form) | Rotala macrandra |
| Duckweed and most floaters (they get aerial CO2) | Sagittaria subulata | Dwarf sag as a carpet, riccia as a lawn |
The floaters row is the cheat code worth noticing: floating plants have access to atmospheric CO2, which is why some (frogbit, water lettuce, red root floaters) grow explosively in tanks where rooted plants crawl. If you want lush without gas, floaters are the low-drama route.
The "wants injection" list is not a morale judgment. Dwarf baby tears in a no-CO2 tank is not a hardy plant being brave; it is a plant slowly starving in the exact currency it needs most. Choose from the first column and the whole CO2 question dissolves.
When CO2 is worth it, and when it is wasted money
Worth it, in rough order of payoff:
- You want a carpet. The classic aquascape lawn species need injection. There is no low-tech dwarf baby tears lawn; there is patience-shaped moss instead.
- You already run a high-output light and plants are starving at midday — pale, leggy, algae on the glass. The light raised the carbon demand; gas is the matching answer.
- You keep red stem plants and want them red, compact, and alive. The strong light those species need pairs with injection by default.
- You want faster growth for aquascaping — trimming, filling in, restructuring on a weeks-not-months timeline.
- Dense stem beds are the design. Injection is the difference between a hedge and a row of sad commas.
Wasted money, equally honestly:
- A 10-gallon with easy plants and the stock light. This is most first tanks, and they do not need gas. A timer and better plant choices beat a cylinder every time.
- Buying gas to fix an algae problem. New tanks bloom; the fix is a shorter photoperiod and patience, not more growth chemistry. Read the light schedule guide before buying anything.
- Buying gas to fix a light-coverage mistake. Gas cannot stretch photons to the dark ends of a tank. Length tables live in best planted tank lights 2026.
- Laminate-grade expectations. If the goal is "some green that lives," the low-tech path delivers it for free.
- No refill path. A cylinder with no shop within a reasonable drive is a paperweight with a gauge. Confirm the refill lane before the regulator, a point the CO2 kit guide makes in detail.
The algae point deserves flattening because it is the beginner trap. CO2 does not cause algae; imbalance causes algae, and the fastest way to create imbalance is high light plus no carbon plan. Adding injection to an already-unbalanced tank is equally capable of making it worse before better. The sequence that works, per both the 2Hr Aquarist and Aquarium Co-Op frameworks: set the photoperiod, pick plants that match your light, let the tank settle, and only then add gas if the plant goals actually require it.
FAQ
Does CO2 hurt fish or shrimp? Not at sane levels. Overdosing can drop dissolved oxygen and pH in ways livestock feels, which is why the standard practice is gas on the light's clock, a drop checker as a trend tool, and a solenoid so nothing runs at night. This page is hardware guidance, not a livestock-health protocol.
Is CO2 worth it on a 10-gallon? Only if the plant list demands it. A 10 with anubias, crypts, and a moss ball has no carbon problem. A 10 with dwarf baby tears and a WRGB-class bar does, and small tanks swing faster when something drifts — the case for starting injection on a small tank is convenience, not safety.
DIY yeast vs pressurized? Yeast bottles are science-fair carbon: real chemistry, wandering rate, no schedule. They can carry easy stems on a budget. They cannot hold the stable level that carpets and reds want, which is the entire point of a regulator and solenoid. The comparison is in the CO2 kit guide.
Does liquid carbon count as CO2? It is a different chemistry — a carbon compound plants can partially use, with mild algae-suppressing side effects — not dissolved CO2. It nudges low-tech tanks along. It does not reproduce injection-level growth, and some plants and inverts react badly to it. Both camps treat it as a middle path, not a substitute.
Can I add CO2 later, or is it a day-one decision? Later, and most people should. Injection is the last variable to add, after light, substrate, and a settled photoperiod. The upgrade path is clean: everything low-tech teaches you (reading plants, water changes, patience) transfers directly.
Will CO2 make my plants grow without good substrate? Root feeders still want a real bed. Gas does not replace soil — see best substrate for a planted tank. Water-column feeders (anubias, java fern) barely care about substrate at all.
Does CO2 change my light schedule? It changes what your light schedule is for. Injected tanks can justify 8–9 hours where low-tech sits at 7–8; the bands and the reasoning live in the photoperiod guide.