I have been in this hobby for almost five years and my tank's lighting has always been a mystery to me. I have read several excellent articles on lighting (see additional reading list at the end of this article), but most of them didn't seem to apply to my specific lighting situation. I have observed many online questions about lighting, and the answers are always incredibly varied because the people who respond typically base their response on their own systems or, worse - from their gut, guessing or repeating what they have read or been told.

Well, for those of you who didn't read my Cirolanid Isopod article, you should know that I don't like gut feelings or guessing when I can get hard data. I soon found Eric Borneman's Lighting Banter and the Problem with Thumb Rules and read Mike Paletta's article that introduced me to the term lux meter. I thought about obtaining one for this voyage of discovery, but lux just sounded too much like lox to me, and that brought back some bad memories.

Photo 1: My Apogee™ quantum meter.

Then I came across Mike Kirda's article that presented some actual data obtained using an Apogee™ quantum meter (see Photo 1). As soon as I heard the term quantum meter, I knew I just had to have one, because "quantum" reminds me of Star Trek, Quantum Leap, quantum mechanics, etc. Mr. Kirda's article does a good job of describing the quantum meter and its limitations; namely, that the Apogee™ meter does not accurately detect the blue spectrum of light, so the PAR values for my VHO lights are most likely underreported by approximately 50%. The Apogee™ meter can measure the PAR of both sunlight and electric lamps, but in both cases the units of measurement are µmol/m-2/s-1. More information on µmol/m-2/s-1 can be found in Kirda's article or in Dr. Joshi's Facts of Light Part II: Photons.

In this article I will focus the results of my findings on some questions I've always wanted to answer with data. If you have been around the hobby for a while I'm sure you've heard many of these questions. How often should I change my metal halide light bulbs? How big is the difference in output between old and new bulbs? Will a better skimmer really help more light reach my corals? This article is the first in a two-part series in which I attempt to answer some of these questions. Part I will focus more on short-term questions, and Part II will focus on answering long-term questions and some questions raised by the data in Part I.

My Setup

I have a 75-gallon main tank (48.5"L x 18.5"W x 21"H; 123.2cm L x 47cm W x 53cm H) with a 29-gallon tank (30" L x 12"W x 18"H; 76cm L x 30cm W x 45cm H) as a sump. While the sand bed's depth varies, it is typically 16-17" (40-43cm) from the water's surface to the top of the sand bed.

My main tank's lighting (see Photo 2) consists of two fixtures:

  • Two 175-watt XM 10,000K single-ended metal halide bulbs powered by two Hamilton magnetic (M57) ballasts. Bulb position and orientation: parallel to front glass and 7.5" above the water's surface.

  • Two 110-watt URI Super Actinic VHO fluorescent lamps in a 46.5" (118.1cm) fixture powered by an Icecap 660 electronic ballast. Bulb position and orientation: parallel to front glass and 8" (20cm) above the water's surface.

Photo 2: Reflector and bulb arrangement in my tank.
   Photo 3: Commercial    Electric's 6500K bulb.

Main Tank Lights:
Two 175-watt XM 10,000K bulbs used for 13 months for 6.5 hours a day, on average.
Two 110-watt URI Super Actinic fluorescent lamps with built-in reflectors used for 13 months for 10 hours per day, on average.

Sump Lighting:
I currently use Commercial Electric's 6500K - 19-watt bulb (Photo 3), typically available at Home Depot.

Methodology and Sensor Holder

As soon as I received my Apogee™ quantum meter, I started taking readings in my tank. I quickly realized that holding the sensor in the tank with my arm would never result in consistent readings. A vertical movement of only 1-2" (2.5-5cm) produced a difference of over 50 µmol/m-2/s-1 in the readings, and a horizontal tilt of 15 or 20 degrees also resulted in a large difference.

Photo 4: Custom sensor holder.

I thought about this problem and collaborated with Nick from Nick's Acrylic Reef to build a sensor holder that I thought would keep the readings consistent (see Photo 4). This would allow me to mount the sensor on a piece of acrylic that would keep the sensor level and minimize the time my arms were in the tank. This was done to prevent oil and shadows from my arm from impacting the sensor readings. I also used a Sharpie™ to mark positions on the holder and the canopy's frame so I could return the sensor to exactly the same location in the tank each time. With this holder, I can confidently state that I was able to consistently measure each position in the tank within 0.5" (1.3cm) horizontally and vertically.

My tank sits within a few inches of a window that receives strong daylight in the early evening, so I ran all tests after it was dark outside and my quantum meter showed 0 µmol/m-2/s-1 when placed on the middle of the windowsill shielded from the light from the tank. All tests were run with the bulbs having been on for at least 60 minutes and having been cleaned with Windex the morning of the tests, unless noted otherwise. This was done to eliminate the variability in PAR caused by salt creep on the bulbs. I feed my tank two to four times a week and to maintain consistency in measurements, I always measured the evening following a day of feeding, typically around 36 hours after feeding.

I carefully planned the locations inside and outside of the tank where I would take measurements. Photo 5 shows all the points inside the tank where measurements were taken, and these points are referred to in the results below. I also took measurements outside of the tank to see the effects of light exiting out of the tank. All measurements, however, were 0-3 µmol/m-2/s-1, so I chose not to include any locations outside the tank in this article.

Photo 5: Data collection points.

I then spent some time collecting data trying to answer my questions. I immediately noticed fluctuations, however, while taking a single reading at the same position, of 30 PAR, which would have made drawing any kind of conclusions from the data difficult. I decided to reinforce the sensor holder by putting two screws on the tank mount bracket and the horizontal mount bracket. This reduced the variability in the readings, but there was still too much fluctuation. I then realized that my Rio SEIO™ powerheads were causing a lot of disturbance on the water's surface. With the SEIO™ powerheads shut off while taking readings, the fluctuations dropped to 0-4 PAR depending on which position I was measuring, which was good enough to get some relatively consistent readings. I observed that fluctuations dropped to zero if the return pump was turned off, but that lowered the water level, thereby changing the PAR readings more than the small fluctuation caused by the return pump. Because I wanted to know what light intensity the corals were receiving, I decided to leave the return pump running instead of lowering the water level.

In the next section of this article I will ask some questions and answer them with the data collected in my tank. Please keep in mind that these answers are specific to my particular lighting setup and methodology. Just because the data indicate a particular answer in my tank, they might indicate something very different in your tank. If you choose to use a sensor holder in your tank, I would be very interested in talking about your results. Some of the answers will help people maximize the utilization of their current lighting, regardless of their light setup.

The Results by Question and Answer

Tests 1 and 2 - Question: Does cleaning the bulbs and center brace really help increase light intensity in the tank?

Test Setup:

Tests 1 and 2: I deliberately let salt spray accumulate on my main tank lights and center brace for two months with no cleaning. Salt spray on the bulbs was minimal as they are 7.5" (19.1cm) and 8" (20cm) above the surface. I first took a set of readings with a salt-crusted center brace and salt-crusted metal halide bulbs after they had been running for one hour. The VHO bulbs were left off for these tests.

Test 1: I shut off the metal halide bulbs and allowed them to cool, cleaned them and turned them back on and let them run for one hour before taking readings again. Note: Both metal halide bulbs were on during these tests, so it is possible that cleaning one of the bulbs could have affected the readings underneath the other bulb. However, this test was done to see the overall effect of cleaning both bulbs at the same time.

Test 2: I then left the metal halide bulbs on and cleaned the center brace. I then took readings underneath the brace to see the difference between the dirty and clean bulbs and center brace.

Answer for Tests 1 and 2: Cleaning definitely did help. The positions on the left side of the tank had a greater increase in intensity than the positions on the right side of the tank. I assume the difference is because the left metal bulb had a greater accumulation of salt spray than the right metal halide bulb. The left metal halide bulb obviously had more salt spray on its glass shield than the right one did and cleaning had a much larger impact on the values recorded (see Graphs 1 and 2). At 6" (15cm) of depth the PAR reading was 12% higher than the right bulb. But what was even more dramatic was the effect of cleaning the center brace (see Table 1).

Graph 1: Test 1 - Sensor readings for positions 3, 13, 23, 33 before and after cleaning both bulbs.

Graph 2: Test 1 - Sensor readings for positions 5, 15, 25, 35 before and after cleaning both bulbs.

Dirty Bulbs and Brace Reading
Clean Bulbs and Brace Reading
Percent Increase
4 (above brace)
< 1%
14 (2 underneath brace)
34 (12 underneath brace)
Table 1: Test 2 - The effects of cleaning the metal halide bulbs and the center brace on PAR readings underneath the brace.

Test 3: I have often wondered what effect salt creep and lamp position would have on my sump's light intensity. I typically positioned my Commercial Electric sump bulb about 5-7" (13-18cm) above the water surface and at an angle because it allowed me to work easily in the sump area.

Answer for Test 3:
To illustrate clearly how the bulb was usually positioned, I set it up on my quarantine tank (see Photo 6).

Photo 6: The sump bulb's typical position.

I had not cleaned my sump bulb in at least three months, so it was approximately 40% covered with salt creep, and some of the reflector's surface was covered in salt creep as well. I took PAR readings ¾" (2cm) away from the bulb, at the water's surface (7" (18cm) away) and 2" (5cm) and 4" (10cm) underneath the water's surface. I then moved the bulb to just 3" (8cm) above the water's surface and directly over the sensor (see Photo 7) and took PAR readings again.

Photo 7: Directly overhead sump bulb position.

I then cleaned the old bulb and reflector with a vinegar and freshwater soak and took PAR readings in both positions again. I then took the same socket and reflector, put in a brand new bulb and let it run for one hour before taking PAR readings in both positions. The readings were very interesting, and I've listed them in Table 2.

 Bulb Age:
6 months
6 months
6 months
6 months
1 hour
1 hour
 Salt Creep:
 ¾" (1.5cm) from bulb
 Water surface
 2" (5cm) below surface
 4" (10cm) below surface
Table 2: (Test 3) Effects of Bulb Position, Age and Salt Creep on PAR Readings.

With the six-month-old light bulb, cleaning the bulb and reflector increased PAR readings by 26-39%. While this is a good gain, by repositioning the bulb directly over the sensor and closer to the water, a gain of 189-441% in PAR readings was recorded. By replacing the old bulb and positioning the new bulb directly overhead, the PAR readings jumped by 410-779%. This table shows that hobbyists should consider placing their sump bulbs optimally to increase their existing lights' efficiency. Since I changed the sump bulb's position, the Chaetomorpha algae's growth rate has been considerably faster. For safety reasons, however, hobbyists should be sure to leave enough room for their bulbs to stay above water in the case of a power outage and a resulting increase in the sump's water level. In addition, a safety chain or zip ties should be used to prevent the light fixture from accidentally falling into the water.

These data suggest that by simply positioning your bulbs properly, and cleaning them and any braces frequently, the amount of light your animals are receiving can be increased for the meager sum of the cost of a paper towel, a smidge of cleaning solution, and your time.

Test 4 - Question: Does MH bulb orientation change the light intensity reaching my corals?

Test Setup: The first set of data was with the point source inside the MH bulb oriented vertically with the "nipple" pointing upwards as shown in Photo 8. Data were collected, and then the MH bulbs were turned off and allowed to cool. After the bulbs had cooled they were oriented so the point source inside them was oriented horizontally as shown in Photo 9. The bulbs were turned back on and allowed to run for 60 minutes before intensity measurements were taken again.

Photo 8: A single-ended bulb in a vertical point source orientation.

Photo 9: A single-ended bulb in a horizontal point source orientation.

Answer: Possibly. Some MH bulbs are designed to produce even light distribution on all sides, but when I ran my tests on the XM 175-watt single-ended bulbs there was a dramatic difference. It should be noted that double-ended MH bulbs are designed to fit in fixtures with one orientation, so this test addresses only single-ended bulbs.

Vertical orientation dramatically decreased light intensity of light in all areas of the tank compared to horizontal orientation. Light intensity in the water dropped from 8.51% to 51.27%, depending upon the sensor's position in the tank. The most dramatic difference was immediately underneath the MH bulbs as shown in Graphs 3 and 4. Positions 35, 36 and 37 under the right MH bulb had an average of 40% more PAR at 12" under the water's surface. The orientation of the point source might have little effect in other bulbs; it could be another design factor (see Harker 2002a and 2002b).

Graph 3: PAR readings directly under the MH bulbs (positions 3, 13, 23, 33).

Graph 4: PAR readings directly under MH bulbs (positions 5, 15, 25, 35).

Test 5 - Question: My tank has been running hot; can I open the hood to reduce heat without losing too much intensity?

Test Setup: I took four PAR readings at each position in the tank without moving the sensor holder. This was done to eliminate any variability caused by sensor placement. Two readings were taken with the hood open and two with the hood closed.

Answer: Opening the hood will help disperse heat in extreme situations but it is only a temporary solution. My canopy is enclosed on both sides and has only a 3" gap in the back with the front able to be flipped open (see Photos 10 and 11). I have always wondered how much light I was losing when I opened the hood because the room gets significantly brighter.

Photo 10: Canopy closed.
Photo 11: Canopy open.

The PAR values did drop with the hood open, but only an average of 4% throughout the tank compared to when the hood was closed. Testing outside the tank with the hood open and closed showed an insignificant amount of PAR lost to the room at the same level as the sandbed inside the tank (0-3 µmol/m-2/s-1). I am curious how much light is lost in aquaria that have only an open top with a hanging pendant.

Test 6 - Question: I have heard that using "egg crate" will focus the light so that more will enter my tank and less will bleed out of the tank into the room. Is this true?

Test Setup: After reading a thread by Anthony Calfo (2006), I became curious as to the effect egg crate would have on PAR readings in my tank. On the day of this test I first conducted tests with no egg crate over the tank. I then placed newly cleaned egg crate over the tank (tapered side up) and retook the readings. In order to move the sensor holder without disturbing the egg crate, the first 1" (2.5cm) at the front of the tank was not covered in egg crate. I recorded PAR readings for no egg crate, egg crate 3" (8cm) from the MH bulbs and 6" (15cm) from the MH bulbs. Positions 1-7 were not recorded due to interference with the egg crate. The florescent bulbs were turned off for this test.

Answer: Directly under the MH bulbs there was little change in intensity. However, the positions not directly underneath the MH bulbs saw significant decreases in light intensity when egg crate was used, up to 52% (see Tables 3 and 4). Overall, the egg crate 3" (8cm) away from the bulbs lead to a smaller decrease than when the egg crate was 6" (15cm) away from the bulbs, but it was still a significant decrease in almost all positions measured. Outside the tank I saw a slight drop in PAR 18" (46cm) from the tank (2 µmol/m-2/s-1 vs. 3 µmol/m-2/s-1 without the egg crate).

Table 3: Percent change in PAR readings from no egg crate to egg crate 6" (15cm) away from MH bulbs.

Table 4: Percent change in PAR readings from no egg crate to egg crate 3" (8cm) away from MH bulbs.

While I did not report the results here, I also tested the effect of the egg crate on just my florescent bulbs. It was a much less dramatic decrease, but still a decrease. I believe this is because my bulbs are 8" (20cm) from the water surface, and a significant amount of their intensity is already lost by the time the light reaches the egg crate. Perhaps if my florescent bulbs were properly positioned (2-3" (5-8cm) above the water surface), then I would have seen an increase as reported by Anthony Calfo. These results demonstrate the importance of bulb distance from the water surface, distance between the bulbs and the egg crate and actually testing changes in your tank instead of just adopting a new way of doing things.

Test 7 - Question: I have been using a lousy skimmer for awhile now; if I start using a much better skimmer, will the increased skimming lead to more light intensity for my corals?

Test Setup: I had been running my system with a TF100A (Top Fathom) skimmer since it was first set up. Sometimes it would skim well, but typically it did a very poor job. I first took a set of readings the night before I replaced this old skimmer with my new ASM G2 skimmer. I then took readings every 2-5 days for two weeks to see what changes occurred in the light intensity.

Answer: This was the most surprising test result, and it goes to show that the data you collect don't always support your expected outcome. Over the two-week testing period I saw no significant difference in the intensity readings in the tank. I fully expected my cleaner water to allow better transmission of light, but the data rejected this hypothesis.

I originally considered leaving the result of this test out of this article, but decided to leave it in to demonstrate how careful experimental design is needed to get meaningful results. I should have defined what a "better" skimmer was before the test began. For example, was a "better" skimmer one that could produce a cup of skimmate per day versus one cup per week? I could have also decided to assign a ranked value for water clarity or yellow tint before the test began and took pictures to demonstrate the change in water color during the two-week period.

Ultimately, this test was unlikely to succeed in showing a difference in PAR readings associated with a "better" skimmer, mainly because I did not figure out exactly how to fine-tune my skimmer until it had been running for over two months. While my skimmer was producing a cup of skimmate every day during the two-week period, it was a relatively thin skimmate that was a light tea color instead of a dark green skimmate that usually indicates better performance. If someone wants to attempt this test with a new skimmer, they should first learn how to fine-tune the skimmer on another tank before beginning the test. This way the two-week testing period could reflect a more accurate comparison between the old skimmer (assumed to be fine-tuned) and the new skimmer.

It does seem logical if more dissolved organics and particulates are removed from the water that light transmission would increase. I would be interested in hearing what other people find if they can replicate this process when replacing their own skimmer.

Test 8 - Question: What are the PAR values of new XM 175-watt 10000K MH bulbs during the 100-hour burn-in period?

Test Setup: I run XM 10000K 175-watt bulbs on my tank and recently ordered replacements for my aging bulbs. I collected data several times for the old bulbs over a period of months. I then removed the old bulbs, put in the new bulbs and let them run for one hour before taking readings. I frequently see websites (Riddle 2006) and threads that say the 100-hour burn-in period for bulbs is important as the bulbs change in intensity and spectrum. I could not measure their spectrum, but I was able to record their PAR intensity changes over the 100-hour burn-in period and compare them to the intensity from my aging bulbs. The VHO bulbs were left off for this test; only the metal halide bulb intensities were measured.

Answer: There was a dramatic increase in the intensity of light coming from the new bulbs compared to the old bulbs (see Graphs 5-8). While there was an initial spike and some fluctuation in intensity, after 10 hours the PAR readings fluctuated within a narrow range.

Graph 5: PAR values for MH bulbs in positions 1-7.

Graph 6: PAR values for MH bulbs in positions 12-17 (position 11 excluded).

Graph 7: PAR values for MH bulbs in positions 22-26 (positions 21, 24 and 27 excluded).

Graph 8: PAR values for MH bulbs in positions 31-37.

Limitations of the Study

The results of this study are limited in several ways. First, I tested only a single type of ballast and MH light bulb for each test. Second, these results reflect a single tank with only a few tested equipment/practice changes. Third, I did not have a spectral analyzer that would have provided additional data about the differences between the new MH bulbs and the aging bulbs.

I would like to encourage hobbyists reading this article to consider borrowing or purchasing a quantum meter to determine how the above questions could be answered in their own tanks. Some of my test results demonstrate that minor changes such as bulb positioning, cleaning and orientation can have dramatic effects on PAR values. If a hobbyist wants to replicate a change mentioned in this article, they are strongly encouraged to test their tank's PAR values, as their results could be very different from mine. Care should be taken when making changes to lighting as it could easily lead to bleaching or stressing of corals.


I now have some data on what happens to my tank's light intensity based on some simple experiments with a quantum meter. I hope some of these data have helped you understand how different positions and practices can dramatically affect the light intensity in our tanks.

I now have a baseline reading from the start of a new set of MH and VHO bulbs. I will report back to you with Part II of "Fun Times with a Quantum Meter" once I've been running the new bulbs for 12-18 months, depending on when the results warrant reporting. I hope to report several other experiments in the future as well.


I want to thank Nick of Nick's Acrylic Reef for helping me with the design of the sensor holder and for making rapid adjustments to the design after my initial tests. I would also like to thank M.A.R.S.H. member Garrick for his "Fun Times" comments that inspired this article's title.

If you have any questions about this article, please visit my author forum on Reef Central.


Internet Sources:

Borneman, Eric H. 1998. Proceedings of the reef aquarium: lighting banter and the problem with thumb rules. Fish Tales, Number 2. America Online Fish and Marinelife Forum Newsletter.

Calfo, A. 2006. Egg-grate/egg-crate/Lighting grid over... Marine Depot Forums.

Harker, R. 2002a. Product Review: Nipple Nonsense. Advanced Aquarist, 1(1).

Harker, R. 2002b. Product Review: Nipple Orientation of Metal Halide Bulbs Revisited. Advanced Aquarist, 1(3).

Kirda, M. 2003. Lighting in Reef Tanks: Some Actual Data. Advanced Aquarist, 2(8).

Paletta, M. unknown date. Lighting a Reef Tank, Part 1. Marine Depot Educational Center. Accessed July 3rd, 2006.

Riddle, D. 2006. Ultraviolet Radiation and Aquarium Lighting. Aquarium Design Website.

Additional Reading:

Joshi, S. 2006a. Facts of Light Part I: What is Light? Reefkeeping Magazine, 5(1).

Joshi, S. 2006b. Facts of Light: Part II: Photons. Reefkeeping Magazine, 5(2).

Joshi, S. 2006c. Facts of Light: Part III: Making Sense of Light Measures. Reefkeeping Magazine, 5(3).

Joshi, S. 2006d. Facts of Light: Part IV: Color Temperature. Reefkeeping Magazine, 5(4).

Joshi, S. Complete Listing of all my Reef Related Articles.

Riddle, D. & Olaizola, M. 2002. Lighting the Reef Aquarium - Spectrum or Intensity? Advanced Aquarist, 1(2).

Reefkeeping Magazine™ Reef Central, LLC-Copyright © 2008

Fun Times with a Quantum Meter, Part I: A Short Term Study of Various Practices' Effects on PAR Values by Brian Plankis -