Key takeaways
- Measure rim to substrate rather than tank height, because a gravel bed removes 5 to 10 cm from the figure that actually matters.
- Intensity falls with the square of distance and then again to absorption, so a 55 cm column needs roughly four times the surface reading of a 25 cm one for the same result at the base.
- Shallow tanks have the opposite problem, since most planted bars over light them and the dimmer becomes a control you use daily.
- Past about 55 cm of water, plan on a pendant, a hanging kit, and low to medium light species unless the budget is genuinely open.
- A manufacturer PAR chart is worth reading only if it states the measuring distance and whether the sensor was in air or underwater.
The listing says the fixture suits tanks up to 120 litres. That is a box size rather than a light specification, and it is why half the planted tanks I see are lit correctly for the wrong shape of water.
Two tanks can both hold 120 litres. One is 100 cm long and 30 cm deep, the other is a 45 cm cube. The same bar is generous over the first and close to useless over the second.
Depth is the dimension that decides this, and it is the one nobody puts in the product title. Here is how to size a fixture from it in about five minutes.
Measure the only dimension that matters
Take a tape measure to the water surface, not the top of the glass. You want the distance from the waterline down to the top of the substrate, which is where your carpet and your crypt crowns actually sit.
A 45 cm tall tank with 7 cm of soil and 3 cm of freeboard is a 35 cm lighting problem, not a 45 cm one. That correction alone moves plenty of tanks down a whole fixture class.
Write the number down in centimetres. Everything else in this article keys off it.
Why depth charges you twice
The first loss is geometric. Light spreading from a source dilutes with the square of distance, so doubling the gap between diode and substrate leaves roughly a quarter of the intensity behind.
The second loss is the water itself absorbing and scattering photons, and it is not even handed. Deep red goes first, which is why the same plant looks different at the base of a tall tank than it does near the surface.
Tannins, suspended fines, and a spotted cover glass each add their own share. The mechanics of that decline are set out properly in how intensity is counted and where it has to be read.
Fixture class by water depth
The table below assumes clear water, a mid range white LED, and a target of 40 PAR at the substrate, which is solid medium light. Scale the surface figures up by about half again if you want a demanding carpet.
| Rim to substrate | Needed just under the surface | Fixture class | Mounting | Realistic ceiling without gas |
|---|---|---|---|---|
| Under 25 cm | 55 to 70 PAR | Entry level planted bar | Resting on the rim | High light reached easily, dim it back |
| 25 to 40 cm | 80 to 120 PAR | Mid range planted bar | Rim, or 10 to 20 cm above | Medium comfortably, high with effort |
| 40 to 55 cm | 130 to 200 PAR | High output bar or small pendant | 20 to 30 cm above the rim | Medium, and high only with injection |
| Over 55 cm | 220 PAR and upward | Pendant on a hanging kit | 30 to 45 cm above the rim | Low to medium unless the budget is open |
Read the shape of that progression rather than the individual figures. Every step down the table roughly doubles what the fixture has to produce for the same result on the substrate.
A rimless tank filled to within 2 cm of the top is a slightly harder lighting problem than a rimmed tank filled 5 cm down, because the extra water sits between the fixture and the plants. Small margins, but they stack up on a tall tank.
Reading a manufacturer chart without being misled
A published PAR chart is the most useful thing a maker can give you, and about half of them are presented in a way that flatters the product. Three questions sort the honest ones out.
- At what distance was each reading taken, and does the chart cover more than one?
- Was the sensor in air or underwater? An aerial quantum sensor reads high when submerged, which is why underwater models carry a correction factor. Apogee document that effect at Apogee Instruments.
- Is the figure from the centre line only, or does the chart show the corners as well?
Fluval publish depth referenced output data for their planted range at Fluval, which is a reasonable model of what a usable chart looks like. If a maker gives you a single number with no distance attached, treat it as advertising.
Shallow tanks, and the opposite problem
Anything under 25 cm of water is easy to light and easy to over light. A modern planted bar sitting on the rim of a 20 cm deep shallow tank can push 90 PAR onto the substrate, which is competition territory.
The result is predictable: a carpet that grows fast for three weeks, then green spot algae on every leaf that stops growing. Buying a dimmable fixture matters more here than buying a powerful one.
My working rule for shallow scapes is to buy one class down from what the listing suggests and run it at 50 to 70 percent. Where that light lands over the footprint matters as much as how much of it there is, which is the subject of how coverage and intensity pull against each other.
Past 55 cm, the decisions get expensive
Tall show tanks, cubes, and 90 cm high display pieces all sit here. A bar resting on the rim wastes most of its output on the top third of the water and starves everything below it.
A pendant hung 30 to 45 cm above the surface delivers a wider, more even column and lets you tune intensity by moving a wire rather than replacing hardware. Expect to pay for it, and expect to run injected gas if you want anything demanding.
The realistic answer for most deep tanks is to plant for the light you can afford. Anubias, Java fern, and Bucephalandra tied to hardscape in the upper half of the scape will look better than a failed carpet, and the full argument sits in what a tall water column does to a lighting plan.
One fixture or two
Depth decides intensity, and length decides how many units you need. A single 60 cm bar centred over a 120 cm tank produces a bright middle and two dim ends, whatever its rating.
On tanks past 100 cm I would rather see two smaller bars spaced at roughly the quarter points than one powerful unit in the centre. Corner readings landing below about half the centre reading are the threshold where plants start leaning inward.
Front to back matters on wide tanks as well. A 60 cm deep footprint under a single narrow bar leaves the front glass shaded, which is exactly where you planted the carpet.
Sizing a fixture in five minutes
Here is the sequence run on a real tank. A 90 by 45 by 45 cm aquarium with 7 cm of aquasoil and the water 3 cm below the rim gives 35 cm from surface to substrate.
The owner wanted a Monte Carlo carpet, which needs somewhere around 45 PAR at the base. From the table that means 90 to 130 PAR just under the surface, putting the tank firmly in mid range planted bar territory rather than the entry level unit already on it.
We fitted a bar with a published chart showing 48 PAR at 40 cm underwater, mounted on the rim, and set it to 80 percent so there was headroom in both directions. Substrate PAR measured 44, the carpet closed in eleven weeks, and the plant demand figures that drove the target sit in the light requirement table for common species.
Buy for the plant you want in two years, then dim to what you need now. Adding a hanging kit and a slider is cheap. Replacing an underpowered fixture eighteen months later is not.
If the maker publishes nothing at all
Budget fixtures frequently list wattage, lumens, and a tank size range, none of which answer the question. Wattage in particular has not tracked output since the LED transition, for reasons worked through in why the old wattage formula stopped predicting anything.
Borrow a meter if you possibly can, because one afternoon of readings settles the tank for its whole life. Aquarium societies commonly own one and lend it out for a deposit.
Failing that, buy dimmable, mount adjustable, and start low. A fixture you can turn down is recoverable from any mistake, while one you cannot turn up is a purchase you make twice.
Questions readers ask about this
Around 50 cm from the waterline to the substrate is where rim mounted bars start losing the argument. Below that you are relying on a narrow hotspot down the centre line and accepting shaded corners. Past 55 cm, a pendant on a hanging kit does the job properly and gives you height adjustment as well.
Rim mounting suits water under about 40 cm deep, where you need every photon you can get. Hanging suits deeper tanks and any scape where even coverage matters more than peak intensity. A hanging kit at $20 to $40 also gives you a continuous intensity control that costs nothing to use.
If the footprint is more than about 45 cm front to back, usually yes. A single narrow bar lights a strip and leaves both the front glass and the back wall dimmer. Two bars set at the quarter points, or one wide angle fixture, will keep corner readings above half the centre value.
No, and running it flat out shortens its life for nothing. A nano bar producing 60 PAR at 15 cm delivers perhaps 12 to 18 PAR at 50 cm through water, which is below what most plants tolerate. Depth needs a different class of fixture, not the same one working harder.
Find the chart row matching your rim to substrate distance rather than your tank height, and check whether the reading was taken underwater. If the chart only quotes 15 cm in air, assume roughly a quarter of that figure at 40 cm of water, then verify with a borrowed meter before planting anything demanding.
