Dana:
Advanced Aquarist is to be commended for attempting to bring to the attention of hobbyists a potentially useful new research technique. Pulsed Amplitude Modulation (PAM) has gradually gained acceptance among marine biologists and is now considered a useful tool for exploring the relationship between reef environmental factors and photosynthesis. There's every reason to think that it can benefit reefkeepers as well. Unfortunately, I’m concerned that the article suffers from several errors and questionable assumptions. The conclusion of the article is particularly disturbing given that it requires several “leaps of faith” in analyzing the study’s data. Since the intensity versus spectrum debate is so important, I hope you’ll address my concerns.
1. The Ocean Optics spectrometer peaks at 4000 counts and clips all signals that exceed this level. It is clear from the graphs that the sensor was too close to the bulbs and that the highest emission lines are clipped. The sensor should have been moved further from the bulbs in order to accurately measure the spectrum of each bulb.
2. The use of colored filters to estimate the proportion of various wavelengths of light is potentially useful, which is why I first introduced the concept to the hobby in the May/June 1997 issue of Aquarium Frontiers. The technique, however, is not without potential pitfalls.
The text notes measuring blue and red wavelengths. There is no mention of a green filter, but the “spectral composition” pie charts show proportions of blue, green, and red. Did you measure just blue and red and calculate green, or were all three bands measured? There is no mention of the brand or model number of filters. Despite your expressed reservations about the filters, the very best Schott RGB filters do a very good job of helping characterize the relative light quality of metal halide bulbs. Which filters did you use?
The two charts suggest that the 4000 K bulb produces about one-quarter blue light while the 12,000 K bulb produces about half blue light. It is not clear from either the charts or text whether total PAR is equal for the two bulbs. In my 1997 article I demonstrated that the German 20,000 K bulb appeared blue compared to a Japanese 6500 K bulb not because the German bulb produced more blue light but instead because it produces less light in the yellow and longer wavelengths. PAR measurements of the two bulbs at some standard distance would have been a useful addition to the data provided. Did you measure the bulbs in this way?
On many occasions you've expressed concern about UV and its impact on coral health. The spectrum graphs suggest a UV spike with the 6000 K bulb. Did you attempt to quantify this spike? Did you make any effort to filter UV during the PAM measurements?
3.You characterize respirometry as, “an inexact science, fraught with all the drawbacks of experiments conducted in...experimental chambers.” This seems a rather harsh description of a methodology that remains a cornerstone of photosynthesis research. In fact, most PAM studies also include respirometer data.
4. Perhaps the most troublesome comments are regarding the actual PAM assumptions and procedures involved in reaching your conclusions. You write, “If one were to measure the fluorescence of this same spot when exposed to different spectral qualities, an estimation can be made of the light source’s ability to promote photosynthesis.” This is a leap worth debating on several levels.
First, you note that the coral is acclimated to “~350 uMols.m2.sec.” (Presumably this is a typographic error. Mols in this context is not plural or capitalized and the phrase is written umol/m^2/sec.) You then note that the three treatments are “46, 85, and 127 mMols m^-2s-1.” I can’t reproduce the actual way that the script was written, but there are multiple problems with the way the units are listed. Assuming you meant umol/m^2/sec, the intensity you are measuring is well below the peak intensity to which the coral had been acclimated.
There’s no mention of attempting to acclimate the coral to these light levels. Apparently you removed the coral from the flow-through tank, placed it in “a shallow basin” and then took your measurements. Studies have shown PAM measurements to be sensitive to temperature. How was a constant temperature maintained in a shallow basin lighted by a metal halide bulb? Were the measurements alternated between bulbs or was intensity increased (or decreased) for a single bulb and then the second bulb evaluated?
Apparently your measurements were made at night. I'm sure you are aware of the literature suggesting that photosynthesis varies in the course of the day. Isn't there a problem looking at photosynthesis during a time when there would normally be no photosynthesis?
You mention using an Apogee sensor to measure PAR in the basin. The sensor has a serious measurement error measuring short wavelength light. Measuring the PAR of a 12,000 K bulb with an Apogee sensor can significantly underestimate total PAR. Since this number is part of the calculation of ETR, an error here is not trivial. Was any correction applied to the Apogee reading?
5. Your conclusion rests on figure 7 showing ETR versus PPFD. It purports to show no difference between the 6000 K and 12,000 K bulbs. There appear to be numerous problems with this graph. You allude to additional “information” not shown on the graph. What was this information?
The graph shows three data points per bulb per light level. Are there any other data points not shown? The three data points per bulb are regressed to produce the curves shown. Three data points are not enough to conclude that the regression curve should be non-linear. You could have just as well used linear regression with only three points. You must have decided a priori that you would generate a non-linear curve. It is also generally dangerous to extrapolate beyond the highest data point when regressing data, particularly when non-linear regression is arbitrarily applied. Since you have, we ought to consider its implication. Accepting that the regression line is accurate, the lines cross at 150 umol/m^2/sec. Since the coral is acclimated to a light level twice the highest shown, one wonders what would have been found at the acclimated light level. Assuming the lines continue as they appear to be headed, the 12,000 K bulb would have a much higher ETR at the acclimated PAR. This would suggest that at acclimated intensities, color does matter. Wouldn't you agree?
6. The Beer et.al article shows yields considerably higher than your measurements for Favia. What do you think caused your yields to be so much lower than in the Beer article?
7. You cite the Kinzie articles, but appear to have overlooked one key conclusion of the authors. Initially the authors concluded that light quality mattered, but then they rethought their conclusions when they realized that the acclimazation treatments had a greater impact on the corals than the actual treatment color. This is analogous to your coral acclimating to natural sunlight (essentially 5000 K light) and then being measured under 12,000 K light. Based on the Kinzie articles, one would hypothesize that the coral would be better adapted to the 6000 K bulb than the 12,000 K bulb. To test this hypothesis, one would have to acclimate the coral to each bulb before measuring. This apparently wasn’t done. Wouldn't it have been better to let the coral acclimate to each bulb before determining photosynthesis rates?
I hope you’ll address the questions I’ve raised. It appears that PAM may ultimately help us answer the questions you raise in the article, but unfortunately as far as I can tell, they can’t be answered with the data you’ve provided us. Unless issues such as acclimation are addressed, the question of light quality versus quantity remains unresolved.
[ February 17, 2002: Message edited by: rharker ]
[ February 17, 2002: Message edited by: rharker ]</p>
Advanced Aquarist is to be commended for attempting to bring to the attention of hobbyists a potentially useful new research technique. Pulsed Amplitude Modulation (PAM) has gradually gained acceptance among marine biologists and is now considered a useful tool for exploring the relationship between reef environmental factors and photosynthesis. There's every reason to think that it can benefit reefkeepers as well. Unfortunately, I’m concerned that the article suffers from several errors and questionable assumptions. The conclusion of the article is particularly disturbing given that it requires several “leaps of faith” in analyzing the study’s data. Since the intensity versus spectrum debate is so important, I hope you’ll address my concerns.
1. The Ocean Optics spectrometer peaks at 4000 counts and clips all signals that exceed this level. It is clear from the graphs that the sensor was too close to the bulbs and that the highest emission lines are clipped. The sensor should have been moved further from the bulbs in order to accurately measure the spectrum of each bulb.
2. The use of colored filters to estimate the proportion of various wavelengths of light is potentially useful, which is why I first introduced the concept to the hobby in the May/June 1997 issue of Aquarium Frontiers. The technique, however, is not without potential pitfalls.
The text notes measuring blue and red wavelengths. There is no mention of a green filter, but the “spectral composition” pie charts show proportions of blue, green, and red. Did you measure just blue and red and calculate green, or were all three bands measured? There is no mention of the brand or model number of filters. Despite your expressed reservations about the filters, the very best Schott RGB filters do a very good job of helping characterize the relative light quality of metal halide bulbs. Which filters did you use?
The two charts suggest that the 4000 K bulb produces about one-quarter blue light while the 12,000 K bulb produces about half blue light. It is not clear from either the charts or text whether total PAR is equal for the two bulbs. In my 1997 article I demonstrated that the German 20,000 K bulb appeared blue compared to a Japanese 6500 K bulb not because the German bulb produced more blue light but instead because it produces less light in the yellow and longer wavelengths. PAR measurements of the two bulbs at some standard distance would have been a useful addition to the data provided. Did you measure the bulbs in this way?
On many occasions you've expressed concern about UV and its impact on coral health. The spectrum graphs suggest a UV spike with the 6000 K bulb. Did you attempt to quantify this spike? Did you make any effort to filter UV during the PAM measurements?
3.You characterize respirometry as, “an inexact science, fraught with all the drawbacks of experiments conducted in...experimental chambers.” This seems a rather harsh description of a methodology that remains a cornerstone of photosynthesis research. In fact, most PAM studies also include respirometer data.
4. Perhaps the most troublesome comments are regarding the actual PAM assumptions and procedures involved in reaching your conclusions. You write, “If one were to measure the fluorescence of this same spot when exposed to different spectral qualities, an estimation can be made of the light source’s ability to promote photosynthesis.” This is a leap worth debating on several levels.
First, you note that the coral is acclimated to “~350 uMols.m2.sec.” (Presumably this is a typographic error. Mols in this context is not plural or capitalized and the phrase is written umol/m^2/sec.) You then note that the three treatments are “46, 85, and 127 mMols m^-2s-1.” I can’t reproduce the actual way that the script was written, but there are multiple problems with the way the units are listed. Assuming you meant umol/m^2/sec, the intensity you are measuring is well below the peak intensity to which the coral had been acclimated.
There’s no mention of attempting to acclimate the coral to these light levels. Apparently you removed the coral from the flow-through tank, placed it in “a shallow basin” and then took your measurements. Studies have shown PAM measurements to be sensitive to temperature. How was a constant temperature maintained in a shallow basin lighted by a metal halide bulb? Were the measurements alternated between bulbs or was intensity increased (or decreased) for a single bulb and then the second bulb evaluated?
Apparently your measurements were made at night. I'm sure you are aware of the literature suggesting that photosynthesis varies in the course of the day. Isn't there a problem looking at photosynthesis during a time when there would normally be no photosynthesis?
You mention using an Apogee sensor to measure PAR in the basin. The sensor has a serious measurement error measuring short wavelength light. Measuring the PAR of a 12,000 K bulb with an Apogee sensor can significantly underestimate total PAR. Since this number is part of the calculation of ETR, an error here is not trivial. Was any correction applied to the Apogee reading?
5. Your conclusion rests on figure 7 showing ETR versus PPFD. It purports to show no difference between the 6000 K and 12,000 K bulbs. There appear to be numerous problems with this graph. You allude to additional “information” not shown on the graph. What was this information?
The graph shows three data points per bulb per light level. Are there any other data points not shown? The three data points per bulb are regressed to produce the curves shown. Three data points are not enough to conclude that the regression curve should be non-linear. You could have just as well used linear regression with only three points. You must have decided a priori that you would generate a non-linear curve. It is also generally dangerous to extrapolate beyond the highest data point when regressing data, particularly when non-linear regression is arbitrarily applied. Since you have, we ought to consider its implication. Accepting that the regression line is accurate, the lines cross at 150 umol/m^2/sec. Since the coral is acclimated to a light level twice the highest shown, one wonders what would have been found at the acclimated light level. Assuming the lines continue as they appear to be headed, the 12,000 K bulb would have a much higher ETR at the acclimated PAR. This would suggest that at acclimated intensities, color does matter. Wouldn't you agree?
6. The Beer et.al article shows yields considerably higher than your measurements for Favia. What do you think caused your yields to be so much lower than in the Beer article?
7. You cite the Kinzie articles, but appear to have overlooked one key conclusion of the authors. Initially the authors concluded that light quality mattered, but then they rethought their conclusions when they realized that the acclimazation treatments had a greater impact on the corals than the actual treatment color. This is analogous to your coral acclimating to natural sunlight (essentially 5000 K light) and then being measured under 12,000 K light. Based on the Kinzie articles, one would hypothesize that the coral would be better adapted to the 6000 K bulb than the 12,000 K bulb. To test this hypothesis, one would have to acclimate the coral to each bulb before measuring. This apparently wasn’t done. Wouldn't it have been better to let the coral acclimate to each bulb before determining photosynthesis rates?
I hope you’ll address the questions I’ve raised. It appears that PAM may ultimately help us answer the questions you raise in the article, but unfortunately as far as I can tell, they can’t be answered with the data you’ve provided us. Unless issues such as acclimation are addressed, the question of light quality versus quantity remains unresolved.
[ February 17, 2002: Message edited by: rharker ]
[ February 17, 2002: Message edited by: rharker ]</p>
