Li-Cor vs Apogee Quantum Meter

[rharker]

Dana,

I would encourage readers to look at the Apogee review I wrote for Advanced Aquarist many years ago. It can be found at www.reefkeepers.info under the lighting section. It will provide additional information not found in your article.

I hope you'll take time to clarify some points of your interesting article:

I was surprised that you used a white bucket rather than a reef tank to compare the sensors. The bright curved sides of translucent white buckets can change the light field in ways quite different from a typical reef tank. How did you settle on using a bucket and what tests did you run to convince yourself that the results would be applicable to a reef tank?

Out of curiosity I did my own experiment. I measured light intensity from the perfect point source (the sun) in the open, and within a white and black bucket. At a time when the sun was overhead, I measured intensity of 1491 uE/m^2/sec with a flat terrestrial sensor. At the same time, my 4pi sensor measured 1490 uE in the black bucket and 3378 uE in the white bucket! That an increase of 127% just by using a white bucket!

Because of the high intensity in the white bucket I couldn't try the Apogee in the open sunlight. Taking the Apogee into the shade, it read 114% higher in the white bucket compared to the black bucket. That compares to just a 77% increase doing the same comparison with a flat terrestrial LiCor sensor. So the Apogee isn't acting like a flat or spherical sensor when it is in the white bucket.

In the article you argue without any supporting evidence that only down-welling light matters, "you'll probably want to obtain a sensor that measures only downwelling light." From a cost saving's standpoint, I would certainly agree. Spherical light sensors are considerably more expensive, but I can't think of another reason. Kirk (whom you cite) along with every other author I'm familiar with stress the importance of measuring the entire light field, not just direct light when measuring underwater light fields.

In my work I've found underwater light fields to be complex and unpredicitable. I have to agree with the scientific studies on the matter. A flat sensor will more often than not under-measure total integrated light. From the data I've collected, I've concluded that a real tank's light field can not be calculated using the inverse square rule, nor can it be characterized by measuring with a flat sensor-unless your reef tank is in a black bucket with black sand!

I've got additional methodological questions. In nearly all of the graphs the Apogee and LiCor numbers are equal at low intensities and then gradually diverge, in some cases dramatically. And these are fairly low light levels. As many tanks have much higher light levels, do you think the difference would be that much greater at more typical light levels? Do you have any idea why the Apogee progressively measures lower levels as the intensity increases? You don't touch on that in your conclusions.

You say the water depth was 18 cm and there are 6 data points on each graph. The X axis isn't labeled, so it isn't clear what is being measured. Can you clarify that? If the X axis is distance, I'd guess that each data point would be (maybe) 2 cm from one-another. Those are terribly small increments to have the lines diverge so quickly. There's no mention of how the distances were measured or controlled. Was the light raised? The bucket lowered? The bucket is a collimator if the light source is high up and less so as the light is lowered, so you are changing other characteristics of the light field as you change the intensity. Did you study that at all?

Your graphs show the AB 10K peaking at 160 uE while the old Coralife 20,000 K peaked at 400 uE. Are you suggesting that the Coralife creates that much more light than the AB bulb?

In your "worse-case" example, your 20,000K measurements in the electric light setting show the Apogee under-reporting light intensity by 23% at the highest intensity. This seems like a large difference and one that refutes the notion that Apogee measurements are interchangable with the LiCor. A hobbist with a lot of actinic light is going to appear to have less light than he really does if measured by the Apogee. Wasn't that the point of the criticisms of the Mike Kirda article?

Richard Harker

 

[Nautilus1]

Seems like Advanced Aquarist will publish anything these days! Quality control with more peer review to accurately decide on what information is valid.

 

[AF Founder]

Seems like Advanced Aquarist will publish anything these days! Quality control with more peer review to accurately decide on what information is valid.

Ad hominem statements are logical fallacies, and generally meaningless. If you have something factual to complain about state it so it can be addressed by Dana.

You should know that Dana and Richard have been going at it for at least 10 years, and that we consistently publish the highest quality material for the reef hobby -- do a search for Joshi, Sprung, Delbeek, Toonen, Hovanec, S. Michaels, Schiemer, etc.

 

[RiddleLabs]

Aloha Richard et al.,

My schedule has been much too busy to allow me to respond to your questions until now. I must agree with Terry Siegel – if there are any factual errors in this article, please state them specifically.
I will categorically state the fallacies in your statements:

Statement by Richard Harker: “In the article you argue without any supporting evidence that only down-welling light matters, "you'll probably want to obtain a sensor that measures only downwelling light." From a cost saving's standpoint, I would certainly agree. Spherical light sensors are considerably more expensive, but I can't think of another reason. Kirk (whom you cite) along with every other author I'm familiar with stress the importance of measuring the entire light field, not just direct light when measuring underwater light fields.”

Reply: This is simply not true. My article lists the Li-Cor website which contains PDF files on their products. Since one considering a purchase of an expensive instrument should investigate such a purchase, I would hope that one would take the time to see Li-Cor’s recommendations. Were you familiar with this site, you would known that Li-Cor states their spherical 4 pi sensor is suited best for phytoplankton work, hence, they state: “(PPFD)… is important, for example, when studying phytoplankton which utilize radiation from all directions for photosynthesis.”

I am astounded that you ‘can’t think of another reason’ for using a 2 pi sensor. You have chosen to ignore comments by Marlin Atkinson (of Hawaii Institute of Marine Biology) and have ignored (or are unaware of) these research papers:

Chalker, B.E. and W.C. Dunlap, 1983. Bathymetric adaptations of reef-building corals at Davies Reef, Great Barrier Reef, Australia. II. Light saturation curves for photosynthesis and respiration. Mar. Biol. Ecol., 73:37-56.

Chalker, B.E., 1977. Daily variation in the calcification capacity of Acropora cervicornis. Proc. 3rd Int. Coral Reef Symp., Miami. II: 417-424.

Chang, S.S., B.B. Prézelin and R.K. Trench, 1983. Mechanisms of photoadaptation in three strains of the symbiotic dinoflagellate Symbiodinium microadriaticum. Mar. Biol. 76:219-229.

And these are just a few references from my collection (the binder with authors’ last names beginning with ‘C’) – there are almost 30 more volumes I could go through!


You’ve also chosen to ignore later works by such noted phycologists such as Falkowski and others. In fact, I have a paper by Bruce Carlson (1999, Organism responses to rapid change: What aquaria tell us about nature. Amer. Zool., 39:44-55) referencing your work with a 2 pi sensor. (Page 49).

In addition, Steve Tyree, in his well researched book ‘Reef Building Stony Corals’ states: “Spherical 4 pi collectors are best for pelagic work while hemispherical 2 pi (are best) for benthic work.” (Page 257).

Statement by Richard Harker: “In your "worse-case" example, your 20,000K measurements in the electric light setting show the Apogee under-reporting light intensity by 23% at the highest intensity.”

Reply: Absolutely untrue. Had you taken the time to contact me and asked for the data, you would have realized that the Apogee underreports by 17% at the upper end of the scale. I can only assume that you garnered an approximation of the % difference by estimating/guessing at the information on one of the article’s graphs. And to make matters worse, you did this twice (once here and again on another list) and both times have stated it as fact. This is beyond irresponsible. You are pleading a case without prior discovery. You owe it to your disciples to be both careful and honest. Frankly, you have been neither.

I stand by all the data in the article. Please state categorically what you know to be untrue in the article and I will respond. Otherwise, I will spend my time more constructively than pointing out flaws in your ‘reasoning.’

Your time is probably best suited in writing an article around a comparison you've conducted instead of attempted to detract from a labor intensive effort. In this manner, you can correct any perceived deficiencies and we will all benefit.

Dana

 

[rharker]

Dana, it is ironic that the editor of Advanced Aquarist would challenge posters to avoid ad hominem comments and you as the author of the article in question would then post a reply filled with them.

The value of a study is only as good as the methodology used to gather the data. If there are errors in the gathering of the data, then the conclusions of a study should be rejected. My questions were directed at your experimental design, chosen methods, and the consistency of your data in regards to your conclusions. The questions are reasonable given the sketchy explanation of why you designed your experiment the way you did. Unfortunately, rather than respond to my questions you simply challenged my right to ask them.

I am astounded that you ‘can’t think of another reason’ for using a 2 pi sensor. You have chosen to ignore comments by Marlin Atkinson (of Hawaii Institute of Marine Biology) and have ignored (or are unaware of) these research papers:
Chalker, B.E. and W.C. Dunlap, 1983.
Chalker, B.E., 1977.
Chang, S.S., B.B. Prézelin and R.K. Trench, 1983.

The newest paper you refer to is 22 years old. Had you cited even older papers, you would have found light levels measured with lux meters. Flat sensors have been around much longer than spherical sensors, they are cheaper, and survive longer in the hands of grad students. It is understandable that there are more articles using them. That doesn't make them the best tool, however.

Had you looked at more recent references, you would have found a growing proportion of authors using spherical sensors. A 2003 article in Coral Reefs by Yates on measuring coral reef community metabolism specifically mentions the use of a 4pi sensor. The authors were studying a coral community--much like we would be doing using a light sensor in a reef tank.

A 2002 Marine Ecology Progress Series article, "How are coral populations structured by light?" used a flat sensor and then took multiple measurements in each location: up, down, and sideways--a poor man's spherical sensor. As you may recall, I recommended that solution to Mike Kirda as a means of compensating for the Apogee sensor's design.

Your reference to Marlin Atkinson seems odd since he co-authored an article in 2001 that measured light with a spherical sensor, "Dependence of calcification on light and carbonate ion concentration," in Marine Ecology Progress Series.

Instruments and techniques evolve over time. What was done twenty plus years ago should not be construed as what we should be doing now. There are more critical issues surrounding your data and conclusions, and that's what we should be discussing.

I stand by all the data in the article. Please state categorically what you know to be untrue in the article and I will respond.

It is difficult to state what is true and untrue when it is difficult to interpret your results. You haven't explained why you chose to use a white bucket, you haven't explained the increased divergence of the sensors with intensity, you haven't explained the undefined X axis on your graphs, and you haven't explained how a Coralife 20,000 K bulb can produce more light than an AB 10,000 K. In short, you haven't addressed any of the questions I asked in the first place.

Rather than challenging my right to question the accuracy of your conclusions, why not address these reasonable questions?

Richard Harker

 

[RiddleLabs]

Richard,

Richard Harker wrote: Rather than challenging my right to question the accuracy of your conclusions, why not address these reasonable questions?

Reply: Richard, my patience with your games is wearing very thin. You know darn good and well that I stated the reasons for using a white bucket in an email to another list we belong to. This was on 7/19/05 and you replied on 7/20/05.

This is what I wrote and you know this very well:

"However, positioning of the sensors would take on even greater importance in an aquarium situation. The white bucket is a 'poor man's' intergrating sphere and helps ensure homogenous light measurements under the conditions of the experiments' situation. The problem with aquaria measurements range from internal light reflections from the glass panes (and especially around corners) to light absorbance and reflection of the substrate and live rock. The problems are compounded by reflection from above by the luminaire. Even with all these challenges, I tried an bare aquaria at first during the experiments, but abandoned that protocol when I found just how much light was exiting the aquarium. The white bucket (intergrating sphere, if you will) greatly increased the amount of PAR available to the sensor and allowed measurements to be taken over a much greater range."

Since both sensors are cosine corrected, it is not unreasonable to expect them to gather light to the extent of their capabilities.

Another error is found in Richard's latest reply. Clearly, a reference to Bruce Carlson's 1999 work is made. The fact that a work is old does not necessarily mean it is outdated. But Richard is the master of 'bait and switch'.

Richard - Your opinions mean little to me. SHOW ME THE DATA!

Dana

 

[rharker]

However, positioning of the sensors would take on even greater importance in an aquarium situation. The problem with aquaria measurements range from internal light reflections from the glass panes (and especially around corners) to light absorbance and reflection of the substrate and live rock. The problems are compounded by reflection from above by the luminaire.

Dana, I agree with you completely. You've made the case for using a spherical sensor in a reef tank, which is why we questioned the use of a flat sensor in Mike Kirda's article. With the ability to integrate light coming from all directions, it can more accurately tell us what light is reaching a given spot in the tank regardless of whether it is direct or reflected light.

Your article had quite a few graphs, but I didn't see any evidence that you tried to analyze your data. Why not "crunch" all the data and see if there are any significant statistical relationships. While I don't have access to the original data, the pattern from graph to graph is quite consistent. Given that consistency, I think there is a very good chance that you would have found statistically valid relationships contrary to your conclusions.

Perhaps you can add the analysis to the article. One very useful relationship to examine would be to look at the correlation between intensity and the difference between the two meters. Because the accuracy of the Apogee declines with intensity in every graph of every bulb, I think you will find a statistically significant relationship between intensity and the inaccuracy of the Apogee meter.

Another interesting relationship to explore is the relationship between accuracy and color temperature. The difference between the LiCor and Apogee meters is small with low Kelvin bulbs and increases with the higher Kelvin bulbs. That suggests you would find a statistical relationship between color temperature and the inaccuracy of the Apogee.

If you want to take it another step, you could run multiple regression and create a formula predicting how much off the Apogee is going to be relative to the LiCor given PAR and color temperature. Now that would be helpful.

If you do the statistical analysis, I think you would find that your data confirms–not refutes what I’ve been saying. The Apogee sensor has significant non-linearities that affect its accuracy, particularly at high PAR and particularly at short wavelengths.

I have a staff that does this kind of analysis and if you want to provide the raw data, I'll be happy to have someone do the analysis for you--for free!

Richard Harker[/quote]

 

[rharker]

Dana, I had one additional question. You observed that the sun setting was closer to the LiCor than the electric setting, but you didn't address the question of why in the article. This multiplier is just an arithmetic term in calculating PAR, so the sun setting simply increases the calculated PAR. Why does the Apogee meter need two calibration settings?

If a hobbyist were to buy an Apogee meter to measure light from fluorescent or metal halide bulbs, a natural assumption would be to buy the meter with the "electric" calibration. Wouldn't it be better then to calculate error based on the electric setting? Why would you ignore those readings and arbitrarily use the "sun" setting just because those readings were closer to the LiCor?

If you're trying to compare two instruments, why wouldn't you use the instruments as they were intended to be used? If there's an electric setting used to measure artificial light, isn't that what we should be comparing the LiCor to?

Richard Harker

 

[RiddleLabs]

Richard,

I have little doubt that, in some cases, you would find a statistically significant difference between measurements made by the Li-Cor and Apogee quantum meters. However, I think you have missed the point.

Let’s look at some realistic scenarios:

Porites coral
Compensation Point: 50 micromol/m2/s
Saturation Point (onset): 150 micromol/m2/s
Photoinhibition (onset): 350 micromol/m2/s

Light source: Iwasaki 6500K metal halide lamp
Apogee meter, ‘sunlight’ setting

At these light intensities, the Apogee meter underreports by at best (or worst, depending upon your perspective) 1%.

Second scenario: Same as above except:
Lighting source: Ushio/AB/Hamilton 10000K metal halide lamp
Apogee meter, ‘electric lamp’ setting (Knowing that the ‘sunlight’ selection gives a better response, I’m unsure of the conditions that would prompt me to use the ‘lamp’ setting. But I’ll humor you).

The Apogee reports conservative measurements of 43 micromol/m2/s
(Li-Cor = 47). At saturation onset, the Apogee says 139, while the Li-Cor equals 159.

As you know, the master of aquatic photosynthesis (John T.O. Kirk) uses a method of determining the onset of photosynthetic saturation that even by his own admission is an estimate (albeit a standardized guess. The same could be applied to photoinhibition).

Knowing that the Apogee meter underreports PPFD in most cases and that we’re dealing with estimated saturation and photoinhibiton levels, should one really be overly concerned with the Apogee’s performance? I suppose the most anal hobbyists would, but I would find their position a bit extreme.

I can’t comment on what my article didn’t contain. I’ve asked you several times during our correspondence for factual errors contained within the article, and you have reported none.

If you’re interested in pursuit of significance or non-significance, I suggest you invest in an Apogee meter and other equipment required to conduct a product comparison. I would view this as a much needed addendum to your 1999 report.

Dana

 

[rharker]

I have little doubt that, in some cases, you would find a statistically significant difference between measurements made by the Li-Cor and Apogee quantum meters. Knowing that the Apogee meter underreports PPFD in most cases and that we’re dealing with estimated saturation and photoinhibiton levels, should one really be overly concerned with the Apogee’s performance? I suppose the most anal hobbyists would, but I would find their position a bit extreme.

If you now acknowledge that there are statistically significant differences between the two meters, why did you write, "measurements were surprisingly close" in your closing and leave it at that? Hobbyists should be given all the facts and analyses. You may consider the differences minor, but why not do the analysis and let the reader decide?

I can’t comment on what my article didn’t contain. I’ve asked you several times during our correspondence for factual errors contained within the article, and you have reported none.

In analyzing a study, identifying missing information is just as important as reviewing the contents of the study. In the case of your comparison, what's missing is at the crux of the article's flaws. So it seems disingenuous to suggest you can't comment on what's missing.

Your factual errors are difficult to identify because of serious gaps in your data. There is no defined X axis, so there's no way of evaluating the reasonableness of your data. You've yet to explain how your old style Coralife 20,000K can produce 150% more light than an AB 10,000K bulb. While you've dismissed my question as irrelevant, it speaks to the reasonableness of your data and is therefore a fair question. Unfortunately, it looks like we're not going to get any additional information from you on these matters. I hope this exchange has atleast touched on some of the issues not addressed in the article and hobbyists have sufficent information to draw their own conclusions about the Apogee meter.

Richard Harker