Coral Coloration & Trace Element Experiment

[taricha]

Let me start off by saying this experiment may end up being a dud - many more paths toward failure than success with this one, but I'll post anyway.
The basis of this experiment is that I have some corals that have really low green fluorescent color, and in the past - I have used Red Sea Trace C (Fe, Mn,... transition metals) to increase the green coloration in some corals.
In at least one case, the coral itself came into the system with good color and lost that color in the system, so I know it's not just the coral individual.

initially a green/red monti digitata (early pictures, left and top - later picture bottom right) lost green skin and went to only brown coloration.
(the tissue loss was sustained during a previous experiment with various sources of high carbon doses that caused tankwide coral stress - the coral recovered and has resumed growing but color is essentially nonexistent.)

So here's the outlines of the experiment and data.
Phase 0: "do nothing" normal tank operation - initial data collection
Phase 1: 6 weeks algal trace element supplementation (red algae - palmaria palmata, and phytoplankton - phytofeast feeding)
Phase 2: 6 weeks of greater than the recommended 10% weekly water changes - 2 to 3 5% water changes per week
Phase 3: 6 weeks of Red Sea Trace A (halogens: I and F) and Trace C (Fe, Mn, transition metals) dosing
Phase 4?: TBD?

Phase 1: The algal feeding phase, I added enough daily phytofeast to be the recommended amount - it turns the water detectably green. It contains at least one strain - synechococcus - that is actually known to be ingested by hard corals commonly in the wild. The red algae is palmaria palmata and is munched by my foxface and urchins. The red algae has replaced 25% of the food protein input. The idea is that perhaps these algal sources are a good choice for traces as they are guaranteed to provide trace elements in biologically relevant amounts and if algal cells are directly ingested by corals, that might be superior to water trace dosing.

Phase 2: Water changes. Water changes fix everything for some people - often for me too. The default in my system previously was an average of around ~1-2% water change per week, so 10-15% per week is an order of magnitude higher. I'll be using Instant Ocean as usual. This represents a pretty low effort way to address trace limitation, if it works.

Phase 3: Red Sea Trace A & C dosing: as I said, in the past - I have visually concluded that dosing Trace C : Iron, Manganese, Cobalt, Copper, Aluminum, Zinc, Chrome and Nickel reversed low green color in some of my corals. When I did that, I had also been dosing Trace A - iodine + halogens - and I know I and F deplete in my system, so I'll address both of those. I'll split the weekly dose to 3x per week.

Phase 4: possibly more detailed element management? possibly a different nutrient strategy (ammonia supplementation)?

If one of the earlier interventions creates good color increase, then the experiment will terminate and later phases won't happen.

Data collection: at the end of each phase the following data will be collected.

Manual Chemistry measurements
pair ICP-OES and ICP-MS
Photos (color-controlled - with color card background) of corals with emphasis on 5 specimens: Green/red Monti digi, Red Monti cap, Green poccilopora, green sinularia, green sarcophyton.
Fluorescent pigment extraction/quantification of samples from each of those corals.

A weakness of the experiment is that there are a number of ways it may be a total bust. Maybe the timing is too short. Maybe the lighting or some other environmental issue is the cause of color loss and trace chemistry doesn't help any of them. Maybe only one or two of the corals are actually trace limited and maybe one of them dies turning this into an n=1 or maybe n=0 experiment.
It could also be a low information experiment if something early "fixes" the problem and later interventions don't get evaluated.
The strongest information-dense result would be if I get to correlate the intervention, ICP data, visual color, and measured fluorescent pigment amount for the nothing+3 interventions.

I'm currently through Phases 0 and 1, and mostly through Phase 2. I'll be sharing data and measurement details later.
I'm open to discussions of what interventions are worth doing in phases 3 (and phase 4 if no nice result by phase 3.)

 

[bubbgee]

Cool. I'm subscribing to this.. Any videos in the future or just pics?

 

[Dragen Fiend]

I know personally I lose green color on one of my zoa's when I stop feeding phyto.

 

[Red_Beard]

Let me start off by saying this experiment may end up being a dud - many more paths toward failure than success with this one, but I'll post anyway.
The basis of this experiment is that I have some corals that have really low green fluorescent color, and in the past - I have used Red Sea Trace C (Fe, Mn,... transition metals) to increase the green coloration in some corals.
In at least one case, the coral itself came into the system with good color and lost that color in the system, so I know it's not just the coral individual.

initially a green/red monti digitata (early pictures, left and top - later picture bottom right) lost green skin and went to only brown coloration.
(the tissue loss was sustained during a previous experiment with various sources of high carbon doses that caused tankwide coral stress - the coral recovered and has resumed growing but color is essentially nonexistent.)

So here's the outlines of the experiment and data.
Phase 0: "do nothing" normal tank operation - initial data collection
Phase 1: 6 weeks algal trace element supplementation (red algae - palmaria palmata, and phytoplankton - phytofeast feeding)
Phase 2: 6 weeks of greater than the recommended 10% weekly water changes - 2 to 3 5% water changes per week
Phase 3: 6 weeks of Red Sea Trace A (halogens: I and F) and Trace C (Fe, Mn, transition metals) dosing
Phase 4?: TBD?

Phase 1: The algal feeding phase, I added enough daily phytofeast to be the recommended amount - it turns the water detectably green. It contains at least one strain - synechococcus - that is actually known to be ingested by hard corals commonly in the wild. The red algae is palmaria palmata and is munched by my foxface and urchins. The red algae has replaced 25% of the food protein input. The idea is that perhaps these algal sources are a good choice for traces as they are guaranteed to provide trace elements in biologically relevant amounts and if algal cells are directly ingested by corals, that might be superior to water trace dosing.

Phase 2: Water changes. Water changes fix everything for some people - often for me too. The default in my system previously was an average of around ~1-2% water change per week, so 10-15% per week is an order of magnitude higher. I'll be using Instant Ocean as usual. This represents a pretty low effort way to address trace limitation, if it works.

Phase 3: Red Sea Trace A & C dosing: as I said, in the past - I have visually concluded that dosing Trace C : Iron, Manganese, Cobalt, Copper, Aluminum, Zinc, Chrome and Nickel reversed low green color in some of my corals. When I did that, I had also been dosing Trace A - iodine + halogens - and I know I and F deplete in my system, so I'll address both of those. I'll split the weekly dose to 3x per week.

Phase 4: possibly more detailed element management? possibly a different nutrient strategy (ammonia supplementation)?

If one of the earlier interventions creates good color increase, then the experiment will terminate and later phases won't happen.

Data collection: at the end of each phase the following data will be collected.

Manual Chemistry measurements
pair ICP-OES and ICP-MS
Photos (color-controlled - with color card background) of corals with emphasis on 5 specimens: Green/red Monti digi, Red Monti cap, Green poccilopora, green sinularia, green sarcophyton.
Fluorescent pigment extraction/quantification of samples from each of those corals.

A weakness of the experiment is that there are a number of ways it may be a total bust. Maybe the timing is too short. Maybe the lighting or some other environmental issue is the cause of color loss and trace chemistry doesn't help any of them. Maybe only one or two of the corals are actually trace limited and maybe one of them dies turning this into an n=1 or maybe n=0 experiment.
It could also be a low information experiment if something early "fixes" the problem and later interventions don't get evaluated.
The strongest information-dense result would be if I get to correlate the intervention, ICP data, visual color, and measured fluorescent pigment amount for the nothing+3 interventions.

I'm currently through Phases 0 and 1, and mostly through Phase 2. I'll be sharing data and measurement details later.
I'm open to discussions of what interventions are worth doing in phases 3 (and phase 4 if no nice result by phase 3.)

Tagging along. I have been wondering some of the very things you are experimenting on.

 

[sixty_reefer]

@taricha are you posting phase 1 later today?

 

[MnFish1]

Let me start off by saying this experiment may end up being a dud - many more paths toward failure than success with this one, but I'll post anyway.
The basis of this experiment is that I have some corals that have really low green fluorescent color, and in the past - I have used Red Sea Trace C (Fe, Mn,... transition metals) to increase the green coloration in some corals.
In at least one case, the coral itself came into the system with good color and lost that color in the system, so I know it's not just the coral individual.

initially a green/red monti digitata (early pictures, left and top - later picture bottom right) lost green skin and went to only brown coloration.
(the tissue loss was sustained during a previous experiment with various sources of high carbon doses that caused tankwide coral stress - the coral recovered and has resumed growing but color is essentially nonexistent.)

So here's the outlines of the experiment and data.
Phase 0: "do nothing" normal tank operation - initial data collection
Phase 1: 6 weeks algal trace element supplementation (red algae - palmaria palmata, and phytoplankton - phytofeast feeding)
Phase 2: 6 weeks of greater than the recommended 10% weekly water changes - 2 to 3 5% water changes per week
Phase 3: 6 weeks of Red Sea Trace A (halogens: I and F) and Trace C (Fe, Mn, transition metals) dosing
Phase 4?: TBD?

Phase 1: The algal feeding phase, I added enough daily phytofeast to be the recommended amount - it turns the water detectably green. It contains at least one strain - synechococcus - that is actually known to be ingested by hard corals commonly in the wild. The red algae is palmaria palmata and is munched by my foxface and urchins. The red algae has replaced 25% of the food protein input. The idea is that perhaps these algal sources are a good choice for traces as they are guaranteed to provide trace elements in biologically relevant amounts and if algal cells are directly ingested by corals, that might be superior to water trace dosing.

Phase 2: Water changes. Water changes fix everything for some people - often for me too. The default in my system previously was an average of around ~1-2% water change per week, so 10-15% per week is an order of magnitude higher. I'll be using Instant Ocean as usual. This represents a pretty low effort way to address trace limitation, if it works.

Phase 3: Red Sea Trace A & C dosing: as I said, in the past - I have visually concluded that dosing Trace C : Iron, Manganese, Cobalt, Copper, Aluminum, Zinc, Chrome and Nickel reversed low green color in some of my corals. When I did that, I had also been dosing Trace A - iodine + halogens - and I know I and F deplete in my system, so I'll address both of those. I'll split the weekly dose to 3x per week.

Phase 4: possibly more detailed element management? possibly a different nutrient strategy (ammonia supplementation)?

If one of the earlier interventions creates good color increase, then the experiment will terminate and later phases won't happen.

Data collection: at the end of each phase the following data will be collected.

Manual Chemistry measurements
pair ICP-OES and ICP-MS
Photos (color-controlled - with color card background) of corals with emphasis on 5 specimens: Green/red Monti digi, Red Monti cap, Green poccilopora, green sinularia, green sarcophyton.
Fluorescent pigment extraction/quantification of samples from each of those corals.

A weakness of the experiment is that there are a number of ways it may be a total bust. Maybe the timing is too short. Maybe the lighting or some other environmental issue is the cause of color loss and trace chemistry doesn't help any of them. Maybe only one or two of the corals are actually trace limited and maybe one of them dies turning this into an n=1 or maybe n=0 experiment.
It could also be a low information experiment if something early "fixes" the problem and later interventions don't get evaluated.
The strongest information-dense result would be if I get to correlate the intervention, ICP data, visual color, and measured fluorescent pigment amount for the nothing+3 interventions.

I'm currently through Phases 0 and 1, and mostly through Phase 2. I'll be sharing data and measurement details later.
I'm open to discussions of what interventions are worth doing in phases 3 (and phase 4 if no nice result by phase 3.)

Question- why do you always use x times the correct dosage or x divided by the recommended dosage. That seems to be a weakness as well

 

[taricha]

Cool. I'm subscribing to this.. Any videos in the future or just pics?

Photos and data.

I know personally I lose green color on one of my zoa's when I stop feeding phyto.

interesting. What phyto do you feed?

@taricha are you posting phase 1 later today?

Not yet. first I'll give some more details about how I'm doing measurements, what the baseline for the system is, and what a "positive" result may look like - based on what I can measure.

Question- why do you always use x times the correct dosage or x divided by the recommended dosage. That seems to be a weakness as well

Yep, I'll try to keep it pretty standard. I don't think "10% water change per week" is anything more than a common guidance or target. When it comes to trace element dosing - I'll stick to manufacturers instructions.

 

[Randy Holmes-Farley]

It will be nice to have any sort of experimental data relating to color. It’s got to be among the most claimed and least proven ideas in reefing.

 

[Dragen Fiend]

Photos and data.

interesting. What phyto do you feed?

Seachem reef phytoplankton.

 

[schooncw]

Seachem reef phytoplankton.

Have you considered or used "live" phyto? Have you considered the differences and made the decision to stick with SeaChem phyto?

 

[Dragen Fiend]

Have you considered or used "live" phyto? Have you considered the differences and made the decision to stick with SeaChem phyto?

I have used phyto feast and OceanMagik. It really comes down to availability. LFS tend to not have it vs running to petco and picking up a bottle of seachem. And no I don't really notice a difference.

 

[taricha]

Trying to quantify coloration in corals for this experiment has led to some interesting observations on the side about how exactly corals generate colors that our eyes see.

I always assumed that the purple mushrooms on the left side of my tank and the right side of my tank were the same. But that seems not to be the case....

Top is how they look under fairly white T5 light and bottom is driven by 450nm LED and photographed with a 90% yellow filter. The left population has significant green fluorescence, while the right population has no green and a bit of red fluorescence. They receive the same light field, so apparently are different communities.

If you look at them very closely under normal tank lighting, you would probably say the left population has a bit more of a blue "sheen" to the purple than the right population does.

Edit: Under my full lighting (T5s, a bit blue) the difference is a bit more noticeable, with the group on the left (top pic) having more of a blue-purple look, and the bottom a more red-purple - pics taken at the same settings/color temp etc.

This is one way that a "blue" color or relative appearance can be generated without any actual "blue" pigments.

 

[Dan_P]

Yep, I'll try to keep it pretty standard. I don't think "10% water change per week" is anything more than a common guidance or target. When it comes to trace element dosing - I'll stick to manufacturers instructions.

Where do manufacturer’s instructions come from I wonder? And in your case, how would the manufacturer know the trace element depletion rate in your system? For that matter how effective is a 10% water change? Consistency is much more important than “how much” when hunting for ghosts :)

 

[taricha]

Ok let's talk about some baseline data of where my system is without/before interventions.

1. Manual Chemistry measurements
2. pair ICP-OES and ICP-MS
3. Photos (color-controlled - with color card background) of corals with emphasis on 5 specimens: Green/red Monti digi, Red Monti cap, Green poccilopora, green sinularia, green sarcophyton.
4. Fluorescent pigment extraction/quantification of samples from each of those corals.
   

I'll leave 3 & 4 the coral observed color and fluorescent pigment extraction quantification to a later post.

Here's baseline chemistry...

I'm generally happy with the agreement between FM, Oceamo (ICP-MS), and my Chem measurements - they are about as close as I could expect from the respective methods. Iodine might be the biggest discrepancy, but it doesn't matter much because they all agree it's very low.
On the right side of the divider, I listed the recommendations from FM, Oceamo, Moonshiners, and Captiv8 (the last 2 also based off the Oceamo ICP-MS report.)
Note that even when based off the same report - the recommendations vary considerably.
Captiv8 largely gives you some NSW measurements as target values, resulting in relatively few recommendations to raise elements - Moonshiners recommends far more additions, but not all common additives are recommended to be dosed.
I color coded red for my water value being too high, blue for too low, green for OK, yellow for element values where there might be a case for dosing, but recommendations disagree.
Four "Trace" elements are in agreement across the board: Cu is high, F, I, and Mn are low.

Let's take Iron as an annoying example of how murky it can be to assess whether to dose an element or not.
Fe according to FM is undetectable, so dosing is recommended by them and anyone else who would base a recommendation off that report. Oceamo ICP-MS finds ~1.5ppb Fe (oceamo calibration data is plenty good to say this is a real detection of Fe) and based on that value, Oceamo, moonshiners, and Captiv8 are all either fine with that level, or would even recommend it should be lower. My intention in Phase 3 is to dose the Fe via Red Sea part C. Red Sea's recommendation is to dose to 150ppb Fe. Literally 100 times the current value and other recommendations (because that's what the Red Sea kit can detect).
At this point the skeptical inner voice should is yelling "What are we even doing here??" the hobby can't decide if we have zero or plenty or if we actually need 100x that much?
Simply put, I plan to do that dosing because in the past in my tank, doing so led to an observed re-gain of green coloration in these type corals. (And hey, maybe depleted Mn was the more important ingredient).

 

[taricha]

1. Manual Chemistry measurements
2. pair ICP-OES and ICP-MS
3. Photos (color-controlled - with color card background) of corals with emphasis on 5 specimens: Green/red Monti digi, Red Monti cap, Green poccilopora, green sinularia, green sarcophyton.
4. Fluorescent pigment extraction/quantification of samples from each of those corals.

Here's the coral color Data:
I'm measuring the color specifically on 5 corals - sinularia, sarcophyton, orange monti cap, pocillopora, and a red/green monti digitata.

Here are the pictures of these corals from Phase 0: baseline.
All photos are shot in the afternoon under my full lighting (T5) with the phone in "Pro mode" at color temp 7500K.

Not all these corals are lacking in color - the sarcophyton is over a foot across and is brightly yellow green, the monti cap is as orange as I would expect it. The monti digi is the one I'm certain has almost none of its intended color (neither red polyps nor green skin) because it lost those colors while in my system. I suspect the sinularia and pocillopora may also be low in fluorescent color - but I do not know that for sure - that may just be their color for these individuals.

To quantify the fluorescent pigment in each coral, after each phase I am doing the following:
Take two cuttings from each coral (except only one of the monti digi) of between 0.10g and 0.50g.
After 5-10 minutes passively drying between two paper towels, the mass of each piece is measured.
Then place each in 5.0mL of distilled water in a centrifuge tube and freeze overnight.
They are then thawed, inverted a few times to mix and refrozen overnight a second time.
Once re-thawed finally, they are inverted to mix and the liquid sample is run through a 0.45um PTFE syringe filter so it's particulate and scattering-free.

The fluorescent pigments being water soluble are easily detected in the water sample after the tissues are disrupted by the 2x freeze/thaw cycle.

I then measure the fluorescent emission from each sample with an excitation light of 405nm and another at 500nm. My spectrophotometer gives no units, and the data is only a relative measure so can't be easily compared to measurements by anyone else - but I'm only interested in comparisons of the same corals over time.

The raw data looks like this:

The Blue data is the extract from the sinularia and the red data is from the orange monti cap. The emission peak at ~500nm from the sinularia is the GFP present in the coral, and the orange fluorescence in the monti cap is the peak visible at ~575nm.The right peaks at 675nm are from chlorophyll A. Chlorophyll A itself is not water soluble, but in dinoflagellates it is in a peridinin-chlorophyllA complex that IS water soluble. So we are looking at symbiont pigment density with that peak.
Based on these fluorescent emission spectra, each of the 5 corals - even the 4 green ones - have slightly different fluorescent proteins than the others.

Then the fluorescent intensity per mass of each piece is plotted and viola...
Phase Zero Fluorescence data:
Fluorescence driven by 405nm LED (red monti cap has none)

Fluorescence driven by 500nm LED (sinularia has none edit: and neither does the pocillopora.)

Symbiont pigment fluorescence - a rough measure of the brown symbiont in the coral.

These numbers are all meaningless at this point - their only value is as a comparison over time from one phase to another.
Note that the replicates can vary quite a lot. The amount of fluorescent protein in a piece of coral can vary massively even from one cut tip piece to another near identical one. Fortunately I'm hoping for very large effects, so if they occur, then signal ought to be far greater than the noise.

 

[BeanAnimal]

I love the idea of a color reference card. I can donate a DGK or similar calibrated waterproof card if you wish.

I am at the gym, but can’t wait to sit down and read this!

 

[taricha]

I love the idea of a color reference card. I can donate a DGK or similar calibrated waterproof card if you wish.

That'd be nice and quite useful.
Currently just using an API master color chart in a ziploc bag weighted down with BBs. A little more widely standard color chart would be cool.

 

[Randy Holmes-Farley]

What sort of fluorescence spectrometer are you using? I'm curious about emission and excitation at the same wavelength.

 

[taricha]

What sort of fluorescence spectrometer are you using? I'm curious about emission and excitation at the same wavelength.

I'm using this one from vernier

But you are right to wonder, in one case I was accidentally measuring garbage.

The sarcophyton detection works without problem. Here's some data from an earlier messing around with spectrally isolating fluorescent pigments, overlaid with the LED emission.

The fluorescent emission for the sarcophyton has a peak around ~502nm (green line) and the "500nm" LED excitation peak is actually about 496nm (black) which has it hitting almost exactly the peak absorbance of the fluorescent pigment (shown in blue). So I get a really strong fluorescent signal from that pigment. I actually have to shorten the collection time for that one or it will nearly max out the meter.

The pocillopora one however is wrong.

The "emission" I was measuring from the 496nm LED was also at exactly 496nm, and was probably just a small amount of scattering in the sample. That pigment can be quantified with the 405nm excitation LED, though - but it's too close to work with the 496nm LED.
The others all have enough space between the LED and the peak fluorescent emission wavelength to work though.

 

[Randy Holmes-Farley]

I'm using this one from vernier

But you are right to wonder, in one case I was accidentally measuring garbage.

The sarcophyton detection works without problem. Here's some data from an earlier messing around with spectrally isolating fluorescent pigments, overlaid with the LED emission.

The fluorescent emission for the sarcophyton has a peak around ~502nm (green line) and the "500nm" LED excitation peak is actually about 496nm (black) which has it hitting almost exactly the peak absorbance of the fluorescent pigment (shown in blue). So I get a really strong fluorescent signal from that pigment. I actually have to shorten the collection time for that one or it will nearly max out the meter.

The pocillopora one however is wrong.

The "emission" I was measuring from the 496nm LED was also at exactly 496nm, and was probably just a small amount of scattering in the sample. That pigment can be quantified with the 405nm excitation LED, though - but it's too close to work with the 496nm LED.
The others all have enough space between the LED and the peak fluorescent emission wavelength to work though.

OK, thanks. I was also confused as the title of the graph above says excitation 500 nm and emission at 500nm, but the legend says excitation says 496 nm, which makes more sense. :)