Interpreting your first Triton Test Results

[stlcard]

Hey guys, new to triton here. Just got the results from my first test. Ive looked around and skimmed a lot of material about triton before starting, either i missed the big article or it doesn't exist, but i need to know how to interpret my test results. I can comprehend the different elements they tested for and the value they found in my water. From there, its just cloudy. I think it would be a great sticky for this board, if someone cleanly explained each column in the test results:

1. what they stand for/mean
2. how to interpret that data
3. and then finally comprehend what to do to correct values that may be out of wack.

Its possible the directions for Triton exist, just maybe they got lost in translation from German to English site versions.

Anyway, long story short. You don't have to give me a full blown run down of A->Z with testing as long as your masters thesis. Just give me some juicy cliff notes on how to understand a test result page, and what we should do about it. Please use my test results (posted below) as an example for us who are new to Triton.

LAST REQUEST: Please do not post within the thread unless you have a legitimate answer or question pertaining to test results pages and "the method" itself. I don't want to gunk this thread up with "nice tank" and "sweet results" comments which just eat up scroll time. If you agree with me in that a clean cut "Directions for understanding Triton Test Results" article or thread would be helpful vote YES above.

Lets get Joe or another Triton Pro in here, and keep the junky thread chatter to a minimum please.

 

[KoleTang]

I'd like to know the accuracy of each test. (ie., Mg measures +/- 0.xx mg/l) Not sure if that information has been made available, but I can't find it.

 

[captainNcredible]

Tagging along

 

[Sangheili]

I'd like to know the accuracy of each test. (ie., Mg measures +/- 0.xx mg/l) Not sure if that information has been made available, but I can't find it.

Accuracies can be found here: http://www.triton-lab.de/fileadmin/user_upload/triton-lab/TRITON_LOD.pdf

Guide to corrections can be found here: http://www.triton-lab.de/fileadmin/triton-lab/TRITON_error_crection_sheet.pdf

Some cliff notes on the various trace elements from Eshan's MACNA presentation:

• Manganese good forgoniopora and algae to grow
• Zinc important for SPS growth - stops burnt tips
• Vanadium needed
• Iodine needed
• Pink in SPS fades with not enough Boron (b-balance)
• Barium bound by montipora digitata
• Strontium needs to be close to 8 for sps and for recovery from fragging
• GFO puts off a little barium and phosphorus
• For example, an aquarium heavily stocked with montipora may require additional zinc or nickel compared to a similar reef aquarium with no montipora

 

[Sangheili]

The results are pretty straight forward but here are some quick notes by column,

• The recommended values (setpoint) are based on research Triton has done of numerous water samples taken from numerous reefs around the world (NWS does not have one exact set of parameter values) and based on what works well in their large display.
• Corrective Dosage is the calculated amount to dose of the Triton Trace Base of that particular element in order to bring it up to the set point. This is an estimate and some elements are consumed very quickly vs. some not at all.
• Maintaining Dosage is the amount of the trace base you would need to dose per day to maintain the setpoint. This is even more of a rough estimate then the Corrective dosage and should be taken very lightly. Best course of action seems to be to make the large corrective dosages and determine your tanks consumption over multiple tests. Once you have a better idea of the consumption you can start dosing more frequently if needed.

I have found the best resources for knowledge about Triton are a) Contacting Triton USA directly b) Triton literature linked in my previous post c) This forum d) The Triton forum on the UK board "Ultimate Reef".

 

[JimWelsh]

Perhaps this will help: My Triton Testing Results: By Randy Holmes-Farley - REEFEDITION

 

[KoleTang]

Accuracies can be found here: http://www.triton-lab.de/fileadmin/user_upload/triton-lab/TRITON_LOD.pdf

Can you explain how to read this? What is LOD, Calibration Max, and Correlation Coefficient?

 

[JimWelsh]

Accuracies can be found here: http://www.triton-lab.de/fileadmin/user_upload/triton-lab/TRITON_LOD.pdf

Actually, measurement uncertainty ("...measures +/- 0.xx mg/l") cannot be determined from this document. It gives a range of correlation coefficients, or R² values, for the calibration curves for the various elements. An R² value of exactly 1 means a perfect fit between the expected (standard) values and the measured values during calibration. The less well the measured values match the expected values, then the lower the R² value will be.

You may notice that several of the calibration curves in the linked document take the form of several data points in the lower left corner of the graph (low range), plus one data point in the far upper right area of the graph (high range). In most cases, the values we care about are going to lie in that low range. The problem is, that one high range data point may tend to make the R² value for the curve be much closer to exactly 1 than it would be if that one high outlier were not included in the calculation.

Let me give a simple example.

If we had the following data points in our curve...

Expected Measured

0.1	0.09
0.2	0.22
0.3	0.27
0.4	0.44
1	0.9

...then the R² value for this data set would be 0.98289 -- not really very good at all (since each measured value was actually +/- 10% of the expected value).

But, if I add just one more data point at a much higher range that does have good accuracy...

Expected Measured

0.1	0.09
0.2	0.22
0.3	0.27
0.4	0.44
1	0.9
10	10

...then that single point causes the R² value to increase to 0.99983!

I would like to see these calibration curves *without* the high range outlier. Even more to the point, though, I'd like to know what the measurement uncertainty is (+/- mg/L, for instance) for the various elements.

 

[JimWelsh]

Can you explain how to read this? What is LOD, Calibration Max, and Correlation Coefficient?

LOD is explained pretty well in this article: Reefs Magazine - Skeptical Reefkeeping XII:

Calibration Max is the maximum value for which a calibration should be considered valid.

Correlation coefficient is a number between 1 (perfect correlation) and -1 (perfectly inverse correlation) that describes the the degree to which the measured values agree with the known actual values of the standards used in the calibration. The closer to 1 it is, the better the accuracy, is the basic idea.