How toxic is ammonia, really?

[Formulator]

In the context of trying to understand why so many people believe Seachem Prime was useful for them in an elevated ammonia situation, against the data in the link below about Prime potentially doing nothing to free ammonia in seawater, the question arises, why?

Does Prime actually "Detoxify" free ammonia, NH3?

This started when @Dan_P was looking at measuring NH3 with seneye and was curious about performance near zero NH3. I suggested trying Prime to artificially zero out the NH3 sensor, and the results were weird... so I checked with my seachem kit. Prime by Seachem is commonly used to treat tap...

www.reef2reef.com

One logical explanation is that folks really do not have a good understanding of how toxic ammonia is. They see 1 or 2 or 5 ppm total ammonia, dose Prime, the fish live, and they declare success.

But how toxic is it really? [note, there are always sublethal effects to toxins, and nowhere in this discussion am I advocating for elevated ammonia, I am focusing solely on the endpoint of death and how it relates to belief in whether Seachem Prime has "worked" in a given scenario]

In other words, what would have happened if they did not use Prime in the identical scenario.

We can use published scientific data on ammonia toxicity to gauge what might have happened.

Of course, every species of fish or invert ay have different ammonia tolerability, and pH also plays a strong role.

But let's explore some data.

Here's a 2019 review article published in INDIAN J. MAR. SCI., VOL. 48, NO. 03, MARCH 2019.

NIScPR Online Periodical Repository: Ammonia toxicity and adaptive response in marine fishes – A review

nopr.niscpr.res.in

The authors are at the University of Antwerp, Biology Department (Belgium) and the Central Marine Fisheries Research Institute in India. Thus, it is not a hobby publication, and they have already collected a fair number of previously published experiments to summarize.

Table 2 is what we want to look at.


Table 2 — Comparative toxicity of ammonia to various marine fish

Species TA-N (mgL-1) UIA-N (mgL-1)
4 -Day LC50 Sea bass 40 1.7
4 -Day LC50 Sea bream 57 2.5
4 -Day LC50 Turbot 59 2.6
4 -Day LC50 Cat fish 45 1.6
4 -Day LC50 Rainbow trout 22 0.3-0.6
8 -Day LC50 Sea bass >22.3 >0.9
20 -Day LC50 Sea bream 15.7 0.89
28 -Day LC50 Turbot 38 1
20-Day EC50 Sea bream 15.7 0.89
28-Day EC50 Turbot 17-19 0.50-0.65
55-Day EC50 Turbot 17-21 0.60-0.75
55-Day EC50 Sea bass 22 0.9
LC50 = Lethal concentration for 50% of the population
EC50 = Concentration reducing growth by 50%

Note the important measure: LC50. That is the concentration where 50% of the fish die in the indicated time.

ALL of the data supports the idea that few of these fish (much less than half) would die in 10-20 ppm total ammonia in 2 days. It takes 4 days in more than twice as much ammonia to kill even half of them.

That concentration is a long, long way above what most reefers would use as a trigger point for using Prime. So when fish survive, it may just be the expected result, whether Prime is added or not.

I don't want to make this initial post too long, so we can follow up with additional papers as they arise showing LC50 data for marine fish in seawater.

But I will note here that these results are typical, not an outlier. This paper shows similar results, where more than 1 ppm FREE ammonia is needed to kill half of a batch of silversides in 96 h. That translates to about 25 ppm total ammonia at pH 8.0.

The acute and chronic toxicity of ammonia to marine fish and a mysid - Archives of Environmental Contamination and Toxicology

The acute toxicity (96-hr LC50) of un-ionized ammonia to mysids (Mysidopsis bahia) and larval inland silversides (Menidia beryllina) was influenced by pH and salinity in a species specific manner. With mysids, NH3 was most toxic at pH 7.0 and less toxic at pH 8.0 and 9.0. In contrast, NH3...

link.springer.com

I’m still skeptical. I think it is more likely that the test kits are not appropriate methods to discern the difference between free ammonia and the detoxified iminium salt. Until I see some GC results with unquestionable specificity for ammonia, I will continue using Prime in my clownfish fry tank.

To understand how an iminium salt might interfere with a salicylate-based ammonia test, let’s break down the steps of the typical salicylate-based ammonia test and hypothesize how an iminium salt could participate in the reactions.

Steps of Salicylate-Based Ammonia Test:

1. Ammonia (NH3) reacts with hypochlorite (OCl−) to form monochloramine (NH2Cl).

2. Monochloramine reacts with sodium salicylate and sodium nitroprusside to form an indosalicylate complex, which produces a green or blue-green color.

Hypothetical Mechanism of Iminium Salt Interference with a Salicylate Reagent-Based Ammonia Test Kit:

Iminium Salt Formation: Suppose the water conditioner forms an iminium salt in the water, such as R-CH=NH2+ (where R is an organic group).

Test Kit Reactions:

1. Reaction with Hypochlorite: The iminium ion could react with hypochlorite (similar to ammonia):

R-CH=NH2+ + OCl− → R-CH=NCl + H2O

In this reaction, the iminium ion (R-CH=NH2+) might react with hypochlorite to form a chloroiminium ion (R-CH=NCl).

2. Intermediate Formation: The chloroiminium ion (R-CH=NCl) or its hydrolysis product might then interact with the test reagents. For instance:

R-CH=NCl + H2O → R-CHO + NH2Cl

This reaction could regenerate monochloramine (NH2Cl) or other reactive intermediates that might proceed through similar pathways as ammonia.

4. Reaction with Salicylate and Nitroprusside:

NH2Cl + salicylate + nitroprusside → indosalicylate complex (colored)

The monochloramine or other nitrogen-containing intermediates from the iminium ion reaction could react with salicylate and nitroprusside to form a colored complex.

5. Color Development: The final colored complex might be similar to or different from the one formed by ammonia, leading to the misinterpretation of the test results as indicating the presence of ammonia.

 

[Randy Holmes-Farley]

I’m still skeptical. I think it is more likely that the test kits are not appropriate methods to discern the difference between free ammonia and the detoxified iminium salt. Until I see some GC results with unquestionable specificity for ammonia, I will continue using Prime in my clownfish fry tank.

To understand how an iminium salt might interfere with a salicylate-based ammonia test, let’s break down the steps of the typical salicylate-based ammonia test and hypothesize how an iminium salt could participate in the reactions.

Steps of Salicylate-Based Ammonia Test:

1. Ammonia (NH3) reacts with hypochlorite (OCl−) to form monochloramine (NH2Cl).

2. Monochloramine reacts with sodium salicylate and sodium nitroprusside to form an indosalicylate complex, which produces a green or blue-green color.

Hypothetical Mechanism of Iminium Salt Interference with a Salicylate Reagent-Based Ammonia Test Kit:

Iminium Salt Formation: Suppose the water conditioner forms an iminium salt in the water, such as R-CH=NH2+ (where R is an organic group).

Test Kit Reactions:

1. Reaction with Hypochlorite: The iminium ion could react with hypochlorite (similar to ammonia):

R-CH=NH2+ + OCl− → R-CH=NCl + H2O

In this reaction, the iminium ion (R-CH=NH2+) might react with hypochlorite to form a chloroiminium ion (R-CH=NCl).

2. Intermediate Formation: The chloroiminium ion (R-CH=NCl) or its hydrolysis product might then interact with the test reagents. For instance:

R-CH=NCl + H2O → R-CHO + NH2Cl

This reaction could regenerate monochloramine (NH2Cl) or other reactive intermediates that might proceed through similar pathways as ammonia.

4. Reaction with Salicylate and Nitroprusside:

NH2Cl + salicylate + nitroprusside → indosalicylate complex (colored)

The monochloramine or other nitrogen-containing intermediates from the iminium ion reaction could react with salicylate and nitroprusside to form a colored complex.

5. Color Development: The final colored complex might be similar to or different from the one formed by ammonia, leading to the misinterpretation of the test results as indicating the presence of ammonia.

You are skeptical of what exactly? Testing if the effect of Seachem Prime by the very method they propose works with it (their test) shows nothing.

You hypothesize a structure but we have tried very hard to understand what chemistry is going on and nothing seems to add up chemically.

 

[Formulator]

You are skeptical of what exactly? Testing if the effect of Seachem Prime by the very method they propose works with it (their test) shows nothing.

You hypothesize a structure but we have tried very hard to understand what chemistry is going on and nothing seems to add up chemically.

I propose that they don’t fully understand how their test kit works with the detoxified ammonia. Conversion to iminium salt as they claim as their detoxification mechanism, could interfere. It just hard to argue with results of so many breeders using the stuff. I understand and think you make a valid argument that ammonia is just not quite as lethal as we often assume, but fry may be an exception and I think very sensitive. I did see the silverside larvae paper, but it does not seem to translate to my experience with clownfish fry.

I don’t know how ethical the experiment would be, but perhaps I will split my next brood into 2 tanks and forego the Prime in one tank. I think you may be surprised by the results. I hypothesize 100% loss in the untreated tank.

In the end we can hypothesize on paper, but I think doing an ammonia-specific test like GC-MS could settle the argument without question. Something just isn’t adding up between the current experimental data and real-world experience, despite your proposition regarding the true toxicity of ammonia in adult fish.

Does anyone have access to a GC-MS system? My lab is too highly regulated to do anything fun with…

 

[Randy Holmes-Farley]

I propose that they don’t fully understand how their test kit works with the detoxified ammonia. Conversion to iminium salt as they claim as their detoxification mechanism, could interfere. It just hard to argue with results of so many breeders using the stuff. I understand and think you make a valid argument that ammonia is just not quite as lethal as we often assume, but fry may be an exception and I think very sensitive. I did see the silverside larvae paper, but it does not seem to translate to my experience with clownfish fry.

I don’t know how ethical the experiment would be, but perhaps I will split my next brood into 2 tanks and forego the Prime in one tank. I think you may be surprised by the results. I hypothesize 100% loss in the untreated tank.

In the end we can hypothesize on paper, but I think doing an ammonia-specific test like GC-MS could settle the argument without question. Something just isn’t adding up between the current experimental data and real-world experience, despite your proposition regarding the true toxicity of ammonia in adult fish.

Does anyone have access to a GC-MS system? My lab is too highly regulated to do anything fun with…

Have you read the very long experimental thread?

I think designing additional experiments is fine, but it’s very complicated to use analytical devices to prove that a weakly bound compound is not just falling apart in the testing, releasing ammonia. Maybe head space gc, but again, it’s not trivial or even direct.

Before getting too wound up in details, it appears you believe the claim of “complexed iminium salt” that Seachem claims.

What structure do you think that actually is, and what compound could form it from ammonia in seawater at pH 8?

 

[Formulator]

Have you read the very long experimental thread?

I think designing additional experiments is fine, but it’s very complicated to use analytical devices to prove that a weakly bound compound is not just falling apart in the testing, releasing ammonia. Maybe head space gc, but again, it’s not trivial or even direct.

Before getting too wound up in details, it appears you believe the claim of “complexed iminium salt” that Seachem claims.

What structure do you think that actually is, and what compound could form it from ammonia in seawater at pH 8?

An interesting possibility would be Glyoxal (OCH-CHO) as the carbonyl source for nucleophilic addition of NH3 with elimination of water. pH 8 is not ideal, but should be basic enough considering the high reactivity of Glyoxal and nucleophylicity of NH3. It is non-toxic and could form a bis-iminium compound with 2 ammonia molecules:

1. OCH-CHO + 2NH3 → 2OHC-CH=NH + H2O
2. OCH-CH=NH + H+ → OCH-CH=NH2+
3. OCH-CH=NH2+ + Cl- → OCH-CH=NH2+(Cl-)

 

[Randy Holmes-Farley]

Glyoxal was found to react with ammonium to form imidazole, imidazole-2-carboxaldehyde, formic acid, N-glyoxal substituted imidazole, and minor products at very low concentrations.

https://www.semanticscholar.org/paper/Glyoxal-in-aqueous-ammonium-sulfate-solutions%3A-and-Yu-Bayer/949ef9cb63fbb6e79ae5774fc0aede6f9d555c11

 

[Formulator]

Glyoxal was found to react with ammonium to form imidazole, imidazole-2-carboxaldehyde, formic acid, N-glyoxal substituted imidazole, and minor products at very low concentrations.

https://www.semanticscholar.org/paper/Glyoxal-in-aqueous-ammonium-sulfate-solutions%3A-and-Yu-Bayer/949ef9cb63fbb6e79ae5774fc0aede6f9d555c11

I need time to read that, but I can see how the imidizole would form rather than proceeding the imine intermediate to imidium salt…. Sulfate in the water maybe a problem. Hmmm

 

[Randy Holmes-Farley]

I need time to read that, but I can see how the imidizole would form rather than proceeding the imine intermediate to imidium salt…. Sulfate in the water maybe a problem. Hmmm

One of the reasons I never liked organic chemistry is that one can draw all sorts of reactions that might happen, but knowing which one does is often unknown without doing experiments.

In any case, remember this product also (or perhaps only) rapidly depletes chlorine, so it likely already has nucleophiles and so something that reacts with nucleophiles might not be stable.

 

[MnFish1]

One of the reasons I never liked organic chemistry is that one can draw all sorts of reactions that might happen, but knowing which one does is often unknown without doing experiments.

In any case, remember this product also (or perhaps only) rapidly depletes chlorine, so it likely already has nucleophiles and so something that reacts with nucleophiles might not be stable.

One of the reasons I never liked organic chemistry is it took hours to study for the tests. However, Without wanting to start another debate - I'm just going to say - there is a difference between an in-vitro test being discussed here, and in vivo detoxification. Whether the Seachem tests work the way they say they should or not - I can say without a doubt (and posted the thread) - that at least one product was a miracle worker with a high nitrite in a (freshwater of course) discus tank.

 

[Formulator]

One of the reasons I never liked organic chemistry is it took hours to study for the tests. However, Without wanting to start another debate - I'm just going to say - there is a difference between an in-vitro test being discussed here, and in vivo detoxification. Whether the Seachem tests work the way they say they should or not - I can say without a doubt (and posted the thread) - that at least one product was a miracle worker with a high nitrite in a (freshwater of course) discus tank.

That’s kind of my hang-up with the conclusion that these products are purely snake oil. I have seen them work (or I thought I did), even if the LD50 is much higher than most of us assumed before the literature review Randy provided. And being a scientist, I don’t like it when the in-vivo and in-vitro tests don’t align. Usually that means the in-vitro test method is flawed or not representative. Sometimes it means there is wishful thinking and a placebo effect, but that’s why we have ex-vivo test methods as well. The last product I worked on was a drug for muscular dystrophy. Clinical results looked great with physical endurance tests, but muscle biopsy of patients showed the drug was not doing anything beneficial.

I need more empirical test results! I will ponder and maybe I can use this as an opportunity to add value to the community.

BTW, I’m one of those weirdo’s who loved organic chem… I actually switched majors from biology to biochemistry and minor in med chem after organic 1 because I enjoyed it so much. LOL

 

[MnFish1]

That’s kind of my hang-up with the conclusion that these products are purely snake oil. I have seen them work (or I thought I did), even if the LD50 is much higher than most of us assumed before the literature review Randy provided. And being a scientist, I don’t like it when the in-vivo and in-vitro tests don’t align. Usually that means the in-vitro test method is flawed or not representative. Sometimes it means there is wishful thinking and a placebo effect, but that’s why we have ex-vivo test methods as well. The last product I worked on was a drug for muscular dystrophy. Clinical results looked great with physical endurance tests, but muscle biopsy of patients showed the drug was not doing anything beneficial.

I need more empirical test results! I will ponder and maybe I can use this as an opportunity to add value to the community.

BTW, I’m one of those weirdo’s who loved organic chem… I actually switched majors from biology to biochemistry and minor in med chem after organic 1 because I enjoyed it so much. LOL

Yea - Im not sure it's snake oil - but the scientists here - have decided it is - due to the testing done - which is fairly rigorous. It seems clear that at minimum, the Seachem claims about their test kits being able to measure ammonia (free) - levels are incorrect. However. IMHO - this does not mean that ammonia is not detoxified in vivo.

 

[Formulator]

Yea - Im not sure it's snake oil - but the scientists here - have decided it is - due to the testing done - which is fairly rigorous. It seems clear that at minimum, the Seachem claims about their test kits being able to measure ammonia (free) - levels are incorrect. However. IMHO - this does not mean that ammonia is not detoxified in vivo.

I certainly don’t want to discredit any of the work already done on the subject. Honestly I still need to read through it all to understand what challenges have already been refuted and what experimental data exists. At the surface level, I am questioning the validity of any experiments done using hobby grade titration kits that may or may not discern the difference between free ammonia and any less toxic form these bottled conditioners may or may not be converting it to.

 

[Formulator]

I just found this interesting paper that seems to conflict with the data generated with the hobby test kits. They used Seachem Prime as a control in their study of using bottled bacteria products to stabilize fish shipments. See figure 3a showing TAN removed by Prime as a function of Prime concentration. Initial concentration of 4.5 mg Nitrogen/L.

The chemical reaction of ammonia removal instantly occurred within 10 min, and no noticeable concentration change was observed after prolonged incubation (1 hr). With increasing the concentration of the chemical ammonia remover, the ammonia removal performance was enhanced (Figure 3a). The highest concentration of water conditioner (0.5%) removed almost 90% of ammonia within 10 min, suggesting the potential effectiveness of the chemical ammonia remover for fish shipping (Figure 3a).

Application of the consortia of nitrifying archaea and bacteria for fish transportation may be beneficial for fish trading and aquaculture

The paper requires a journal subscription, but PM me if you need a copy of the PDF. Snip below contains the relevant part. It is nice that there is no conflict of interest, it is from a very reputable source, and funded by an NSF grant…

 

[Randy Holmes-Farley]

I just found this interesting paper that seems to conflict with the data generated with the hobby test kits. They used Seachem Prime as a control in their study of using bottled bacteria products to stabilize fish shipments. See figure 3a showing TAN removed by Prime as a function of Prime concentration. Initial concentration of 4.5 mg Nitrogen/L.

The chemical reaction of ammonia removal instantly occurred within 10 min, and no noticeable concentration change was observed after prolonged incubation (1 hr). With increasing the concentration of the chemical ammonia remover, the ammonia removal performance was enhanced (Figure 3a). The highest concentration of water conditioner (0.5%) removed almost 90% of ammonia within 10 min, suggesting the potential effectiveness of the chemical ammonia remover for fish shipping (Figure 3a).

Application of the consortia of nitrifying archaea and bacteria for fish transportation may be beneficial for fish trading and aquaculture

The paper requires a journal subscription, but PM me if you need a copy of the PDF. Snip below contains the relevant part. It is nice that there is no conflict of interest, it is from a very reputable source, and funded by an NSF grant…

In Figure 3 A, isn’t ammonia rising with more ammonia remover added?

You presume all the tests that were run in the associated thread were wrong due to interference, but this test must be right? (Why?)

 

[Formulator]

In Figure 3 A, isn’t ammonia rising with more ammonia remover added?

I thought that at first but reading the methods, they plotted amount removed on the y axis out of a starting concentration of 4.5 mg N/L.

You presume all the tests that were run in the associated thread were wrong due to interference, but this test must be right? (Why?)

I’m not sure yet. I’m not presuming the tests in the associated thread were wrong, but I am certainly questioning them. Thats the whole premise of my argument here. My hypothesis is that the work done so far by members here has all been done using hobby grade kits that can’t discern ammonia from whatever reaction product Prime makes. Yes, I see 3 or 4 different kits/films have been used, but who knows if those are even different methods or not. The paper came out of an accredited university with grant funding from the NSF. I have a hard time believing the results would be unreliable. To me it is more likely the shade-tree chemistry done with hobby kits does not have the specificity required to test Prime’s mechanism of action. I’m going to get some salicylate reagents from Hach and try to replicate the findings in the paper. I have a UV-Vis spec in the garage.

Do you know what reagents the hobby kits are using? I know the salicylate reagent-based method used in the paper uses some hazardous chemicals, so perhaps not the same as the hobby kits.

 

[Malcontent]

This paper was already discussed...

 

[Randy Holmes-Farley]

Another comment on that paper (which I have not seen in full.

Assuming they labeled the graph wrong, and it should be ammonia removed, The doses of chemical remover they need are more than 0.1% to remove 2 ppm ammonia.

The dose that Seachem recommends is:

Use 1 capful (5 mL) for each 200 L (50 US gallons) of new water.

That dose is 0.0025%.

At the Seachem recommended dose, there is nearly no effect.

 

[Randy Holmes-Farley]

This paper was already discussed...

Thank you for the reminder.

 

[Formulator]

Another comment on that paper (which I have not seen in full.

Assuming they labeled the graph wrong, and it should be ammonia removed, The doses of chemical remover they need are more than 0.1% to remove 2 ppm ammonia.

The dose that Seachem recommends is:

Use 1 capful (5 mL) for each 200 L (50 US gallons) of new water.

That dose is 0.0025%.

At the Seachem recommended dose, there is nearly no effect.

I still think it is a piece of evidence in support of my hypothesis that the product of the Prime detoxification reaction is being picked up as ammonia with the hobby test kits. In the previous discussion of the paper, which was brief, it was acknowledged that the salicylate method is the only one in the literature that is picking up the effect of Prime.

And, to be fair, I think Prime instructs to use a 5x dose for ammonia higher than 2 ppm. I also just found in the Seachem website product FAQ - “A Nessler based kit will not read ammonia properly if you are using Prime®... it will look "off scale", sort of a muddy brown (incidentally a Nessler kit will not work with any other products similar to Prime®). A salicylate based kit can be used, but with caution. Under the conditions of a salicylate kit the ammonia-Prime® complex will be broken down eventually giving a false reading of ammonia (same as with other products like Prime®), so the key with a salicylate kit is to take the reading right away.“

 

[Randy Holmes-Farley]

I still think it is a piece of evidence in support of my hypothesis that the product of the Prime detoxification reaction is being picked up as ammonia with the hobby test kits. In the previous discussion of the paper, which was brief, it was acknowledged that the salicylate method is the only one in the literature that is picking up the effect of Prime.

And, to be fair, I think Prime instructs to use a 5x dose for ammonia higher than 2 ppm. I also just found in the Seachem website product FAQ - “A Nessler based kit will not read ammonia properly if you are using Prime®... it will look "off scale", sort of a muddy brown (incidentally a Nessler kit will not work with any other products similar to Prime®). A salicylate based kit can be used, but with caution. Under the conditions of a salicylate kit the ammonia-Prime® complex will be broken down eventually giving a false reading of ammonia (same as with other products like Prime®), so the key with a salicylate kit is to take the reading right away.“

All of us in the Prime thread are well aware of the testing limitations. I wrote about this many years ago. Seachem gives directions on how to test. Following that exact protocol with their tests show no effect.