Coral Coloration & Trace Element Experiment

[taricha]

Oh, they use edta iron? That would reduce mineral surface binding.

Seems so.

 

[Dan_P]

The write up of this thread is next going to be on Phase 2 - Water Changes Fix Everything, but I'm gonna jump to the future for some thoughts here.

In Phase 3, I'm currently dosing Red Sea Trace Part A: I, F, Br and Part C: Fe, Mn, other light transition metals.
Depending on which of the optional dosing methods you use, you might end up dosing very different amounts.
Dose by Calcium demand is one way: 1mL /100L of each product per 20ppm of Ca consumed. For sanity - using kalk, work it out from alk testing to give you a stable kalk dose.

The second way is test and dose: "4 day dosage"
"...test the levels of the Color Elements and supplement as necessary to achieve the optimal levels. Ensure that all of the Trace-Colors™ Elements are at optimal values and run the aquarium for 4 days at a stable salinity (compensate for evaporation daily) without adding any supplements. At the end of the 4 days test the Trace-Colors™ Elements and calculate the
“4 day dosage” of each supplement to replenish back to the optimal levels. Add the “4
day dosage” to the system. Divide this “4 day dosage” by 4 and use as the daily dosage
for the next week."


The optimal values recommended by Red Sea are Iodine = 0.060ppm (fine) and Fe = 150ppb (100x higher than other recommendations for Fe).
After a two week ramp up period, I dosed the optimal levels to look at what 4-day consumption might look like.

Here's iodine over the first 2 days...

And here's Iron for the first 2 days...

(error bars shown on Fe test because the error is large enough to limit measurement at some point.)

Anyway, to my eye - the iodine looks mostly like fast but plausible uptake - I have a couple of gorgonians and some macroalgae in the sump (caulerpa). They've been in a system with depleted iodine for a long time - makes sense they'd have an appetite for it.

The Iron though, looks sketchy - that initial drop of nearly ~100ppb of Fe in ~6 hours just looks like mechanical loss of Iron from the water. Sorta makes me think of how some people do CaCO3 powder "coral snow" to bind and scrub out a bunch of stuff from the water - I wonder if this dose might be doing the same thing with Fe - precipitating a bunch of "Iron Snow". I think I'm likelier to ignore the day 0-1 behavior and take the day 1-4 change in Iron to calculate a more sane daily rate. I think that'll give me a dosage much closer to the "Ca demand" method.
The "4 day dose method" as written would result in me pouring tons of Fe in and possibly precipitating out all sorts of stuff. Not comfortable with that - feels like it might be doing a bunch of stuff that I don't intend it to.

The iron depletion curve looks like mine when I dose iron (II) gluconate.

 

[Dan_P]

Here are iron depletion rates from a study I did this year with Seachem Flourish iron (II) gluconate. Initial concentrations of around 500 ppb in Instant Ocean and aquarium water were measured with the Hanna Checker. I measured iron in unfiltered and 0.45 μ filtered samples.

First plot iron in IO, unfiltered sample. Iron concentration rapidly drops to and holds at about 75% of the original dose. Second plot, filtered samples, shows the amount of filterable iron increases (measured concentration decreases) over 24 hours. This has implications for iron concentrations measured by ICP if samples are filtered.

In the study done in aquarium water, the unfiltered sample results show a different trend from IO (see third plot). The rapid drop is gone possibly because of DOC. Filtered samples of the aquarium water (fourth plot) showed a similar increase in filterable iron (orange plot) as with IO (blue). Apparently, DOC does not effect on the filterable species. The “+PO4” means the presence of PO4 initially thought to play a role in iron concentration (it still might, but my results did not justify rejecting the null hypothesis).

 

[Randy Holmes-Farley]

Interesting. How much phosphate in that experiment?

Could the raw IO have enough bacteria to take up the iron?
Or is it just unstable towards precipitation?

I don't expect you have an answer, but if you do, that's great!

 

[taricha]

First plot iron in IO, unfiltered sample. Iron concentration rapidly drops to and holds at about 75% of the original dose. Second plot, filtered samples, shows the amount of filterable iron increases (measured concentration decreases) over 24 hours.

can you comment more on what might be happening in graphs 1 and 2 (I understand I'm requesting speculation) with the unfiltered sample stabilizing at ~75%, and the filtered sample dropping to ~10-15% over the same time.
And by filtered, I presume you mean that you have a bulk sample with the Fe added and "filtered" means that you draw out and filter a test sample at test time then run the hanna test on it.

 

[Dan_P]

Interesting. How much phosphate in that experiment?

Could the raw IO have enough bacteria to take up the iron?
Or is it just unstable towards precipitation?

I don't expect you have an answer, but if you do, that's great!

Phosphate was 0.4 ppm.

Sorry no answers, but great questions for follow up experiments. A study in a sterile filtered solution would eliminate bacteria and other particles, though aquarium water with lots of bacteria and POC seems to keep iron concentration up.

I was unable to detect a precipitate but I think we are talking about micron size particles and not enough to cause haziness. A very long light path might be useful.

I think the thing to do now is to get @taricha interested

 

[Dan_P]

can you comment more on what might be happening in graphs 1 and 2 (I understand I'm requesting speculation) with the unfiltered sample stabilizing at ~75%, and the filtered sample dropping to ~10-15% over the same time.
And by filtered, I presume you mean that you have a bulk sample with the Fe added and "filtered" means that you draw out and filter a test sample at test time then run the hanna test on it.

Sorry about my communication skills. I was just turning in when I saw your lastest post.

Test results from samples that are not prefiltered stabilize high, but when samples are 0.45 μ filtered before being tested, the measured iron is lower and continues to go down over time.

I guess this means the iron species being filtered off is detectable by the iron test. This happens with our phosphate test as well when aragonite dust (presumably) gets into the test sample and jacks up the phosphate test result.

 

[taricha]

I guess this means the iron species being filtered off is detectable by the iron test.

So the filtration is either removing Fe particulates that would otherwise stay in the water, or it's removing some other particulate that Fe in the water likes to attach to.

BTW, I remember us discussing this data - I thought the stable level after the quick drop was weird. I didn't expect the stable level to show up in my reef tank data, but here we are.

 

[Randy Holmes-Farley]

I doubt this is a critical issue, but when focusing on organics (which may bind iron) it is a known issue that some organics (such as a protein) will stick to syringe filters even if they are dissolved organics.

The IO was normal IO, or RC? The latter has metal chelators but I don’t think the former does.

 

[Dan_P]

So the filtration is either removing Fe particulates that would otherwise stay in the water, or it's removing some other particulate that Fe in the water likes to attach to.

BTW, I remember us discussing this data - I thought the stable level after the quick drop was weird. I didn't expect the stable level to show up in my reef tank data, but here we are.

Yeah,the high stable level is suspicious. It was repeatable at the time but it doesn’t fit how I think “chemistry works”. Do you see ORP providing further insight into what is going on? Maybe this can’t be repeated now and we can all move on

 

[Dan_P]

I doubt this is a critical issue, but when focusing on organics (which may bind iron) it is a known issue that some organics (such as a protein) will stick to syringe filters even if they are dissolved organics.

The IO was normal IO, or RC? The latter has metal chelators but I don’t think the former does.

Normal IO.

Ooo, good idea on filter material vs protein binding. I think that I was using a low protein binding material, but still a good place to investigate.

You and @taricha are making me feel like relooking at these experiments again.

 

[taricha]

Do you see ORP providing further insight into what is going on?

Side by side ORP of reef water with Fe2+ vs Fe 3+ would be measurably different, but I don't see how that would help clarify things.

Ooo, good idea on filter material vs protein binding. I think that I was using a low protein binding material, but still a good place to investigate.

You and @taricha are making me feel like relooking at these experiments again.

I kept the water samples for the Fe decrease data, and it wouldn't be hard to re-check the samples to see how filterable vs unfilterable Fe compares and has changed over time.

 

[taricha]

This has implications for iron concentrations measured by ICP if samples are filtered.

Also - assuming I find a filtered vs unfiltered difference - FM tests are cheap and I wouldn't mind tweaking their protocol and sending an unfiltered replicate to pair with the one that's filtered as instructed. (FM and Oceamo both have you filter the water samples.)

 

[Dan_P]

Side by side ORP of reef water with Fe2+ vs Fe 3+ would be measurably different, but I don't see how that would help clarify things.

I kept the water samples for the Fe decrease data, and it wouldn't be hard to re-check the samples to see how filterable vs unfilterable Fe compares and has changed over time.

I am very, very temped to relook at iron depletion rates again.

 

[Dan_P]

Side by side ORP of reef water with Fe2+ vs Fe 3+ would be measurably different, but I don't see how that would help clarify things.

I kept the water samples for the Fe decrease data, and it wouldn't be hard to re-check the samples to see how filterable vs unfilterable Fe compares and has changed over time.

Had a thought. If the Flourish iron gluconate is a mixture of free and chelated iron (II), maybe that would explain the weird data. I am back in the game. I have both the old bottle and a new bottle of Flourish iron and ferrous sulfate coming next week. I’ll start lab work in a week or so.

 

[taricha]

Here's the 4 day Iron depletion data. I still like the idea of using the day 1-4 data and ignoring the day 0-1 data for calculating my tank's consumption rate. That'd come a lot closer to the "dose by Calcium" recommendation anyway.

Had a thought. If the Flourish iron gluconate is a mixture of free and chelated iron (II), maybe that would explain the weird data. I am back in the game.

Dan, this nugget is for you... I used a cellulose acetate 0.22um syringe filters.
Left is the initial spiked water that I had stored - retested 4 days later, tested it filtered and unfiltered.
The right two pairs are replicates of filtered and unfiltered tankwater taken from the tank on day 4.

It seems very tempting to conclude conveniently that the large fraction of Fe that becomes filterable from the sample is the same large fraction that is rapidly lost from the reef water. What remains detectable in the water 4 days later is entirely un-filterable.

There's no point in me sending filtered vs unfiltered samples for ICP, because at the stable levels that I would send, there's seemingly no filterable Fe fraction.

 

[Randy Holmes-Farley]

Had a thought. If the Flourish iron gluconate is a mixture of free and chelated iron (II), maybe that would explain the weird data. I am back in the game. I have both the old bottle and a new bottle of Flourish iron and ferrous sulfate coming next week. I’ll start lab work in a week or so.

That’s plausible. Old ferrous gluconate might even be susceptible to bacterial degradation in the bottle if nothing in it prevents bacterial growth.

 

[Dan_P]

Here's the 4 day Iron depletion data. I still like the idea of using the day 1-4 data and ignoring the day 0-1 data for calculating my tank's consumption rate. That'd come a lot closer to the "dose by Calcium" recommendation anyway.

Dan, this nugget is for you... I used a cellulose acetate 0.22um syringe filters.
Left is the initial spiked water that I had stored - retested 4 days later, tested it filtered and unfiltered.
The right two pairs are replicates of filtered and unfiltered tankwater taken from the tank on day 4.

It seems very tempting to conclude conveniently that the large fraction of Fe that becomes filterable from the sample is the same large fraction that is rapidly lost from the reef water. What remains detectable in the water 4 days later is entirely un-filterable.

There's no point in me sending filtered vs unfiltered samples for ICP, because at the stable levels that I would send, there's seemingly no filterable Fe fraction.

Good and good. This will help me design my experiments and with luck my weird data goes up in a puff of smoke. I hope so anyway. I don’t need another rabbit hole just now. Thanks!

 

[Dan_P]

Here's the 4 day Iron depletion data. I still like the idea of using the day 1-4 data and ignoring the day 0-1 data for calculating my tank's consumption rate. That'd come a lot closer to the "dose by Calcium" recommendation anyway.

Dan, this nugget is for you... I used a cellulose acetate 0.22um syringe filters.
Left is the initial spiked water that I had stored - retested 4 days later, tested it filtered and unfiltered.
The right two pairs are replicates of filtered and unfiltered tankwater taken from the tank on day 4.

It seems very tempting to conclude conveniently that the large fraction of Fe that becomes filterable from the sample is the same large fraction that is rapidly lost from the reef water. What remains detectable in the water 4 days later is entirely un-filterable.

There's no point in me sending filtered vs unfiltered samples for ICP, because at the stable levels that I would send, there's seemingly no filterable Fe fraction.

One more thought. In my hands, Hanna Fe values around 20 ppb = 0 ppb. The 26 ppb reported on the plots is it the Hanna value or difference from baseline?

 

[taricha]

One more thought. In my hands, Hanna Fe values around 20 ppb = 0 ppb. The 26 ppb reported on the plots is it the Hanna value or difference from baseline?

All the values I used are corrected for the hanna baseline (I got around ~17ppb baseline for my water)
So the 26,27 are raw values of 43,44 on the checker.