Marine Aquarium Lighting

[zanzimog]

So you want to start a marine tank, and maybe have a few corals or macroalgae. Unlike fish, most corals and plants need light that is sufficiently intense and of the right spectrum. How to get the right kind of light can be mysterious and daunting, when faced with all of the acronyms like PC, VHO, T5, as well as the specs like Kelvin value, lumens, PAR, and the like.

The goal of this article is to give an overview of the technology, and hopefully give a novice an idea of what a "6700K PC light" really means. Translating this into what fixture is best for what kind of livestock is another can of worms, and will wait for another day.

I. What do the organisms need?

Photosynthesis is essentially the process of light smacking into pigment molecules and imparting energy into them that is then used to make complex molecules. For this to work, the light has to be from the right part of the spectrum and enough photons need to smack into enough molecules to provide enough energy to keep the organism going. In other words, you need the right spectrum and intensity to get the job done.

Spectrum
The main pigment molecules used by plants and corals' photosynthetic algal symbionts absorb maximally in the 400-550 nm (blue) and 620-700 nm (red) areas of the spectrum. The upshot is that you want a lot of blue and some red, but other wavelengths are not used very efficiently, and UV can be downright harmful.

Another issue affecting the choice of a spectrum is the effect of water on light. Because water absorbs the red end of the spectrum more than blue, white sunlight becomes increasingly blue with depth. Therefore reef tanks look more natural, and the inhabitants may be happier, with light that is blue-white.

One further consideration is that many corals fluoresce green or red when illuminated with blue light, so aquarists try to maximize this fluorescence by enhancing the blue end of the spectrum. One very common way to do this is to supplement daylight with "actinic" lights, which put out a more-or-less pure peak of blue light that excites the corals' pigments efficiently. The blue light from actinics also contributes to photosynthesis.

So how can you tell what the spectrum of your bulbs is? Many manufacturers put a plot of the spectral output on their websites or the packaging of the lamps. That's fairly useful as a starter.

Color Temperature
Another measure of spectral output is color temperature, also referred to as the K value or degrees Kelvin. It is based on the spectral output of an object (in principle an ideal emitter called a "blackbody") when heated to a given temperature. For reference, the Sun is glowing at about 6000 degrees Kelvin, so its output is about 6000 K. Lower temperatures are more red, with the spectrum shifting to blue-white and blue as the temperature increases.

For the reef aquarium, color temperatures range from 6500 K, which is considered "daylight," to 20000 K, which is a very deep blue.

While this measure works very well for incandescent lights, there are some serious limitations with the use of color temperature for fluorescent and metal halide systems. Because these systems depend on phosphors emitting at specific wavelengths (see below), rather than putting out a broad spectral peak, it is difficult, and in some cases impossible to assign credible color temperatures to them. Actininc lamps cannot be assigned a color temperature at all, because they emit at a single spectral peak.

Nonetheless, for the time being, color temperature is one of the best ways of judging what the output of a lamp will look like, and whether it will suit your needs.

One other measure of spectrum is the color rendering index, or CRI. However, this relates to how the human eye perceives light, and is not useful for judging light for a reef tank.

Intensity
In addition to the spctral quality of the light, you need to know the intensity. There are many measures, which have varying ease and utility. People often use watts per gallon (wpg) as a rule of thumb for gauging the ability of a system to support certain organisms. There is some utility to wpg, but it is extremely limited. How much usable light an oranism gets from a watt of electricity can vary tremendously, based on the type of light, the ballast, the color temperature of the lamps, the reflector, the clarity of the water, the depth of the tank, and so on. One might be able to use wpg as a very rough estimate, but it simply can't be used to determine suitability for a given species.

Then there are lumens and lux. Lumens are units of the amount of visible light emitted by an oject, while the amount of light falling on a surface is measured in lux. The problem with either of these measures is that they are biased toward what humans can see, which is different from what corals need.

One of the more useful measures is photosynthetically available radiation (PAR). This is the amount of light falling on a particular surface area, between the wavelengths of 400-700 nm. It is actually a pretty decent approximation of what corals can use, although a PAR meter can be fooled if a lamp puts out a lot of light in the middle yellow-green part of the spectrum. PAR meters are becoming increasingly available to hobbyists, so it is not out of the question to measure the PAR at various places in your tank.

As I mentioned, PAR does not correspond exactly to the regions of the spectrum that photopigments absorb the best, the photsynthetically usable (PUR). At present, measuring PUR is still out of reach of most hobbyists.


II. Lighting Systems.

There are three fundamental ways of turning electricity into light:

Incandescence
Incandescent light is the old fashioned methods of using the glow from a hot filament. This is incredibly inefficient, producing more heat than light, and is not generally used.

LEDs
Light emitting diodes (LEDs) make use of glowing semiconductors. Extremely efficient and long-lived, and finally making a serious entry into the reef aquarium market. I have zero experience with these, but this writeup by Dana Riddle has convinced me that LEDs are finally more than a kiss and a promise: LED review. Although the output is not quite as strong as the most efficient halides, they have two major selling points. First, they produce a lot less heat than metal halides, and can be placed closer to the water. Second, they should maintain their output upwards of 10 years, saving tremendously on bulb replacement. When one considers that lamp replacement for a 48” MH/T5 canopy costs about $175 per year, the $2300 price tag starts sounding better.

LEDs have made a solid inroad in the form of "moonlights." These small lights are designed to come on during the nighttime, to provide an opportunity for nocturnal viewing, and to simulate lunar cycles. I have a few concerns about these units (note: editorial rant coming on). First, having logged many hours of night diving, I can tell you it’s freakin’ dark down there, even with a full moon and dark-adapted eyes. The LEDs are much brighter than moonlight, in a region of the spectrum your fishes’ eyes are very sensitive to. The other concern is that these units are touted to promote reproduction of your corals. In the unlikely event that this is true, spawning events are not necessarily good things in small glass boxes. At best, it is potentially polluting, and at worst some coral gametes are poisonous. Despite my misgivings, these lights are very popular and I have heard no complaints. I would be inclined to use them sparingly, though.

Fluorescence.
This involves pushing electricity through a gas and making it glow. Almost all current aquarium lighting systems make use of this, including fluorescent and metal halide systems. The idea is that the glowing vapor (usually mercury) gives off photons of a particular wavelength. The tube containing the mercury vapor is coated with materials (phosphors) that emit light of desirable wavelengths when the photons from the mercury hit them. Aquarium lamps use a combination of phosphors that combine to put out the desired spectrum.

As I mentioned above, the way that fluorescence works means that the output is made up of discrete peaks, rather than broad spectra. Daylight lamps (6700-10000 Kelvin) use a combination of phosphors that put out strong peaks in the red and blue, with a few little peaks in between, that combine in our eyes to look white. The most common way of increasing color temperature is to reduce the amount of red, rather than increasing the blue. One important consequence is that higher color temperture bulbs almost invariably put out less PAR. In other words, if you want growth, go with low Kelvin rated lamps.

There are two basic types of lighting systems using fluorescence, "fluorescent" and metal halide. Both systems require a ballast to generate enough current to start the gas glowing ("firing" the lamp), and a regulated power supply to keep the process going in a controlled way. Ballasts can be either magnetic or electronic, with electronic ballasts weighing less, using less energy, running cooler, and costing more. In most cases, the ballast is separate from the bulb, either mounted in the fixture, or located at some distance (for MH). However, screw-in compact fluorescent lamps have their own ballasts, and are therefore "self-ballasted."

Fluorescent Lights
There is a bewildering array of fluorescent systems available. The terms T5, T8, T12, etc, describe the diameter of the lamp in 1/8ths of an inch. So T12s are 1.5" across, while T5s are 5/8 of an inch. Here are some general classes.

Normal Output (NO): This is the old standard, seen in either T12 or T8. The output is not very intense, so they are not used much in reef tanks.

High Output (HO): The HO T12 is not a lot better than the old T12, so it's not seen much. However, HO T5 systems are available that can compete favorably with metal halide setups in terms of PAR per watt. The skinny little lamps often have individual reflectors, increasing their efficiency, and T5s continue to increase in popularity.

Very High Output (VHO): T12 VHOs are no more efficient than NO lamps, but are driven at a higher wattage to provide a lot more intensity. Not all that common as a standalone system, but VHO actinics remain very popular because they are said to produce the most intense fluorescence from corals.

Power Compact (PC): Also known as compact fluorescent, the tubes are usually T6s, and come in a wide range of sizes, intensity and spectra. For the moment, PC lights, either in dedicated hoods or as DIY systems assembled from screw-in versions, remain the economical powerhouse of the marine aquarium. The efficiency per watt is generally somewhere between NO and metal halide.

Metal Halides
These are in a class by themselves for a few reasons. They remain the most intense, efficient way of turning watts into PAR, and so are considered the system for demanding corals or seagrasses. Further, because they are similar to point sources, unlike fluorescent tubes, they generate more of a shimmering effect when they shine through the surface of the tank.

Halides tend not to be the first lighting system a hobbyist purchases. MH lights tend to be expensive, especially if they are bought in ready-made hoods. Another important issue is heat. Simply put, a lamp pulling that much wattage gets really hot. The canopy needs to be designed to dissipate the heat, and the lamp needs to be high enough to avoid transferring too much heat to the tank.

MH lamps come in two basic styles, single ended, which look like big lightbulbs (also called a mogul base), and double ended, which are much smaller (in part because they do no have an outer, UV-protective shell). Despite a lot of discussion, I have yet to see data indicating that one is truly more efficient than another.

Photoperiod
Because days and nights are about 12 hours long in the tropics, that’s a good starting point. Most marine setups have two or more sets of lights, one actinic and one daylight, and it is common to run the actinics for the full 12 hours with the daylights on 6-10 hours in the middle. MH lights should be on for no more than 8 hours, and in some cases considerably less, because they use a lot of electricity, produce considerable heat, and because the corals’ zooxanthellae will generally be saturated by that time.

Lamp Replacement
Over time, phosphors start to give out, and lamps will lose intensity at the desired wavelengths. That means you will see less growth of desirable organisms, and it's possible that pests that make better use of the degraded spectrum will spread. So, to maintain the optimal spectrum and intensity all lamps need to be replaced regularly. VHOs last little more than 6 months, while I have seen figures between 6 and 12 months for PC lamps. Most MH lamps lose most of their punch between 10-12 months and should be replaced at least once a year.

Links of potential interest:
Joe Burger's Comparison of MH PAR values
Sanjay Joshi's Comparison of MH spectra
AquaBotanic's Comparison of PAR from NO, VHO, PC and MH lamps
Nuts and Bolts of Metal Halide Lamps and Ballasts

Please post questions, comments and corrections.

[Lobo.]

what does it mean when a light is self-balasted?
how long does it take before you need to replace a light? and why do you have to replace them if they are still glowing?
what are the benefits of night (or moon) lights?
how many hours of light per day is advised?

i thought that I would make you buisy

[zanzimog]

Better?

Gave me a chance to do a little editorializing. One of the benefits of writing these things

[joephys]

Originally Posted by zanzimog

...
For the reef aquarium, color temperatures range from 6500 K, which is considered "daylight," to 20000 K, which is a very deep blue.
...

I have heard that different color tems are used to simulate various deapths. For example, 6500K is daylight, 10,000K is daylight at a deapth of 40 feet and 20,000K is daylight at 80 feet deep. I just made up the deapths so it could be off, but some along those general lines. It makes sense, but are the different colors used for this purpose or something else? Using 10,000K and actinic, I don't see a point for 20,000K bulbs unless that is the reason.

[zanzimog]

I have heard that different color tems are used to simulate various deapths.... Using 10,000K and actinic, I don't see a point for 20,000K bulbs unless that is the reason.

Yes and no. Some use 20000 Ks to simulate a deep reef, but others simply like the blue look and the way the corals fluoresce under high color temp lamps. One guy I know uses 20000 Ks and VHO actinics, because he likes the look so much.

It's really a matter of mixing and matching to suit your needs. I use high PAR 10000 K halides plus T5 actinics to give a look around 14000 K. You could simplify things by just using a pair of 14000 K halides, but you would lose the PAR of the 10000 Ks as well as the light from the actinics.

Best thing to do is look at the tanks in the LFS and local reefers' houses to see what you like the look of, and what gives good growth.

If you haven't checked it out, have a look at this page on Joe Burger's web site. It shows what the output of various lamps and ballasts look like on the same tank.

[forestal]

nice article Dave