the following article was written by regani on the aquariumlife forum (See this article on AquariumLife)

Introduction

Ammonia levels in aquaria and ammonia toxicity to fish is something that often comes up in discussions in internet forums, especially in the context of cycling a new tank, but also as the mysterious “ammonia spike” that seems to appear out of nowhere and kill off the inhabitants of a fish tank that until then seemed perfectly happy.
Despite the fact that ammonia seems such a big trouble maker (or maybe because of it) there are many misconceptions and plenty of wrong information about what ammonia is, where it comes from, and how it affects the fish we are keeping.

In this article I will try to give a bit of a background about ammonia in general, summarize the scientific facts about ammonia in an aquatic environment, and what is currently known about ammonia toxicity in fish.
I have tried to write in a way that is (hopefully) easy to understand for the layperson, but in the interest of avoiding ambiguity I have used scientific formulae and terminology where necessary.
In the reference section at the end I have listed the sources I have used to compile this article; some of the scientific articles will not be directly available (many scientific journals operate on a subscription basis) but may be accessed via some university libraries for anyone who is interested in more details.

The basics apply for both freshwater and marine tanks but because of the much higher salt content and the generally higher pH in the ocean, marine fish have developed some different strategies to deal with ammonia and have somewhat different sensitivities. For those interested I have included a link to a document dealing with ammonia in a marine environment in the reference section.

Where does ammonia come from?

Ammonia (NH3) is made up from nitrogen (N) and hydrogen atoms (H) and in its free state is a gas. Free ammonia gas is quite rare in nature as it reacts readily with water or any even slightly acidic compound to form ammonium salts (more about that in the section about ammonia and pH). Ammonia in free and bound form comes from natural sources but also plays an important role in some industrial processes, especially in the production of fertilizers. For industrial purposes it produced on a scale of about 150 million tonnes per year using the nitrogen gas in the atmosphere (78% of the air we breathe is nitrogen gas) in what it is called the Haber-Bosch process. 
It can also be found in some minerals such as sal-ammoniac (ammonium chloride) or mascagnite (ammonium sulfate), both water soluble salts, or sometimes also as small impurity in some insoluble minerals.

The two major natural sources for ammonia are volcanic eruptions and the decomposition of organic material such as plant matter. As the former generally only plays a minor role in the life of the average aquarist, I will focus on the latter.
The nitrogen bound in organic matter comes mainly from the protein content but can also be found in small amounts in some carbohydrates and lipids. The average protein contains about 15% nitrogen by weight, which is quite a lot.
Let’s just do a quick estimation what that actually means for an aquarium: good quality fish foods will contain 40-50% protein; let’s take a standard 4ft tank that holds about 250L of water; if the tank is fully stocked with good filtration a pinch of food for one feed will be approximately 1g; to make our math easier, let’s assume we have 50% protein content, this means that with each feed we introduce about 0.075g of nitrogen - enough to get the ammonia levels to about 0.35ppm, that is 0.7ppm for two feeds, and a good 1ppm from 3 feeds a day. In a 2ft tank (holding 75L of water) the same amount of food will mean 1.2ppm, 2.2ppm, and 3.3ppm for one, two or three feeds a day, respectively.

Of course there are other factors that come into play as well. Contributing to the ammonia load is any other decomposing organic matter such as dead plants, dead bacteria, or, heaven forbid, a dead fish. Living fish will also excrete ammonia, even if no food is given (see below for details). Some plants can directly absorb some ammonia from the water, but the real reason of course that we are able to keep fish at all are all those nice bacteria that convert the ammonia to nitrate, rendering it (more or less) harmless.

Basic fish biology – digestion and metabolism

I have mentioned that the ammonia in an aquarium comes from the fish food and that fish will also generate ammonia without any food added to the tank. To understand how that happens we need to have a closer look at how the cells in a fish (and almost any other living organism) actually convert food into ‘energy’ that can then be used to grow or move or do a multitude of other things.
The major components in fish food are proteins, fats, and carbohydrates, together with a whole host of other things like vitamins, minerals, etc. Carbohydrates and fats are the major energy sources in food and the proteins are a source of amino acids, which are used to build new ‘fish-specific’ proteins, e.g. in muscle fibres. But any amino acids that are not immediately needed are also converted into energy. 

So, how does that work? It is a fairly complicated process overall, but can be simplified to a few core actions. To start with the food will need to be broken down into small enough ‘bits’ that can then be transported around the body to give each cell what it needs. That is basically two things: material to build ‘stuff’ and energy that drives the machinery that builds ‘stuff’. Although there are many different types of cells in a body, they all accept the same basic building materials and a universal energy unit, so the most efficient way for a body to operate is to break down the food into those building blocks and universal energy units.
The proteins in the food contain the same basic building blocks (amino acids) that are needed to build new proteins in the body of the fish so it’s easy to recycle them to build new proteins by breaking them down into those individual amino acids. Fats and carbohydrates are made up from carbon chains of different size and shapes and are broken down into small units containing only two carbon atoms; these can then be used to either build other fats or carbohydrates (e.g. for energy storage) or they can be converted into universal energy units (called ATP) that can be transported to wherever they are needed. The same will happen to all amino acids that are not immediately needed to build proteins and they are broken down into the same basic units to provide energy. The only problem here is that the amino acids also contain some nitrogen that needs to be removed to allow that. This causes a dilemma as by removing that nitrogen from an amino acid we are generating ammonia! 
This whole process is going on all the time, even when there is no food available; as the fish metabolism just keeps on going – otherwise the fish would die – there is constant need to generate energy and after depleting easily available energy sources the body starts to break down muscle tissue to supply the required energy, and muscle is largely made up out of – guess what – protein! 

As ammonia has the unpleasant side effect of being fairly toxic, different organisms have developed different strategies to neutralize the ammonia generated by breaking down amino acids derived from proteins. In many mammals, including us, the ammonia is converted into urea via some biochemical trickery. Urea is far less toxic than ammonia, so it is safe enough to have it around inside the body and it gets excreted via the kidneys as one of the components of urine. This does cost the body a little extra energy, but unfortunately we can’t just breathe out ammonia as we do carbon dioxide (CO2), so that little extra energy to make ammonia safe is a necessary investment.
The reason we can breathe out CO2 but not ammonia is their different solubility in water. While CO2 is not very soluble in water, can be moved around in concentrations that don’t cause major problems and can be made to leave the water again quite easily, ammonia loooves the water and can easily be dissolved in water up to concentrations that actually dissolve tissue. And once it’s in the water it is not so easy to get it out again. You can boil water to remove ammonia, but who would want to have to run their lungs at 100°C???

What does pH have to do with all this?

After we have done a bit of biology to understand where ammonia comes from we need to do a short excursion into chemistry to understand how it behaves and how pH and ammonia toxicity are related. However, before we move onto ammonia in particular, I’d like to give just a brief summary of what pH actually means and how it works and will try not to get too technical.
Water is truly special stuff and has some very peculiar abilities; if it wasn’t so common it would have to be classed as an ‘exotic’ molecule because of the way it behaves. One of the unusual things it can do is to fall apart to a small degree for a very short period of time! Fortunately this only happens to a very small number of molecules at a time – you wouldn’t want all the water in your aquarium to disintegrate suddenly!
What actually happens is that a water molecule falls apart forming a proton with a positive charge and a hydroxide ion with a negative charge, or written in chemists’ shorthand: 

H2O <<<<--->H+ + OH-

I used the strange arrow to indicate that the equilibrium is far to the left hand side of the equation, i.e. that there is only a small proportion of free hydroxide and protons and that the vast majority of the water exists as H2O. 
(If we wanted to be precise we would have to talk about hydroxonium ions H3O+ instead of protons, as free protons don’t exist in water, but for our purposes and to keep things simple, the good old proton H+ will do.)
In pure water, as there is nothing else around but water, there will always be the same number of both ions. If you make some assumptions and apply some mathematics you end up with some numbers such as the amount of protons in one liter is 1x10-7 Mole (and, no, that is not the cute black little burrowing mammal). We don’t need to concern ourselves too much with the details, what it boils down to is that in pure water the pH=7 by definition. 
When acids are present (such as hydrochloric acid) we have more H+ than OH- and the number for the pH goes down, when bases are present (such as caustic soda) there are more OH- than H+ and the pH goes up. Under normal conditions in water the pH ranges from 0 to 14 and for fishkeeping we only need to concern ourselves with pH values between about 3.5 and 10 (roughly the range in which fish are found in nature). 

Just because we will need it later to get some numbers I will quickly introduce the mathematical equations that we need to calculate pH and some related values. There are two key equations that we will need. One is the Henderson-Hasselbach equation, which allows us to calculate the pH of any given acid or base solution:

pH = pKa + log { [A-]/[HA] }

Here [HA] and [A-] are the concentration of the acid (HA) and what is called the ‘corresponding base’ (A-) and pKa is basically the strength of the acid we are looking at. (In chemistry the square brackets often mean a concentration.) We don’t really need to concern ourselves with details here, but if anyone wants to know more, there are plenty of sources to explain how the above equation is derived, what it means and the limitations it has.
The other equation is the equation for the Gibbs free energy of the dissociation of the acid (the dissolving of the acid in water):

ΔG = 2.303 RT * pKa

Here ΔG is the free energy of the reaction, pKa we already know, T is the temperature, and R is the gas constant. More about this a bit later, but you can see that temperature is a factor in one of the equations, so you may start to suspect that pH is possibly temperature dependent – and you would be right to suspect that. This is one of the reasons why temperature has an effect on the toxicity of ammonia as well; again, more about this later.

Ammonia and water – the chemist’s perspective

So, what does that mean for ammonia? Despite the fact that ammonia (NH3) contains hydrogen atoms similar to hydrochloric acid (HCl) it dissolves in water in a different way. Hydrochloric acid dissolves in water to form protons (H+) and chloride (Cl-):

HCl (gas) (+ H2O) --->>> H+ + Cl-

You may have notice that I used a different arrow compared to the equation for pure water. With hydrochloric acid, because it is a ‘strong’ acid, once it dissolves in water into H+ and Cl-, that’s basically it; it does not easily go back the other way. This means there is plenty of additional H+ around and the solution is acidic, i.e. has a pH<7.
When ammonia dissolves in water the equation looks like this:

NH3 (gas) + H2O <--->> NH4+ + OH-

Two things are different this time; one is that the ammonia dissolving in water does not form protons but hydroxide (OH-) and ammonium (NH4+) ions, and the arrow is different again indicating that most of the ammonia dissolves to form hydroxide and ammonium, but that some also goes back the other way so that there always is some free ammonia present as well. The fact that hydroxide ions are formed instead of protons means that ammonia acts as a base, not an acid, so the pH goes up when ammonia is dissolved in water and it goes up more if we dissolve more ammonia. In the equation above it doesn’t go all the way to ammonium because ammonia is not really a ‘strong’ base, more like a ‘medium’ one.
The laws of chemistry now stipulate that if by dissolving and changing the concentration of ammonia (or other gaseous acids or bases) we change the pH of the water, the reverse must also be true, namely that if we change the pH of the solution we can change the amount of free ammonia. So what happens if we add some acid or base? Let’s look at the relevant equations:

NH3 (gas) + H2O <--->> NH4+ + OH-

H+ + OH- <--->>>> H2O

If we add acid (H+) to the ammonia solution, the protons will react with the hydroxides to form water. Because our ammonia is a law-abiding citizen in the world of chemistry it is desperately trying to follow the above equation and keep up the concentration of hydroxide to compensate for the acid ‘stealing’ the hydroxides. This it can only do by reacting more free ammonia with more water to form more hydroxide - and ammonium. This way the lower the pH (more H+) the more ammonium (NH4+) and the less free ammonia (NH3) there is, although the overall number of both combined stays the same. 
The opposite is true when we add a base: with more hydroxide present good old ammonia tries to right things again by backpedaling on its reaction with water and form less hydroxide, which means turning ammonium (NH4+) back into free ammonia (NH3). This means that at higher pH there is more free ammonia (NH3) present, although the overall number of both combined again stays the same.
With the help of the mathematical equations from the previous section we can now calculate how much free ammonia and ammonium we have at any given pH and we can do this for different temperatures. Without going into details, if you substitute parts of one equation into the other, apply some mathematical tricks, make some adjustments so that we can use temperature in °C (not in Kelvin as required in the original equation), and write it down in a particular way, you end up with:

[NH3] = [NH3 + NH4+]/(1+10^{(0.0902-pH) + (2730/(273.2 + T))})

So, if we measure the concentration of total ammonia ([NH3 + NH4+]), the pH, and the temperature we can calculate how much free ammonia we have in our water sample.
To save you the trouble of having to do it yourself, I have put some results in the table below as examples. 

In the top row is the total ammonia as can be measured with a standard test kit, below it are a few different temperatures, and on the left the pH of the solution. The values in the table show the amount of actual NH3 in the water, which is the most important in terms of toxicity.

It is clear from the table that pH has a significant impact on the presence of free NH3 and also that temperature plays an important role. For example at 1ppm total ammonia ([NH3 + NH4+]) the amount of free NH3 at pH 5 is negligible, but if we raise the pH to 8 at 18°C the free NH3 rises to 0.098ppm and at the same pH but at 30°C this more than doubles to 0.203ppm free ammonia!

Ammonia toxicity in fish

So far we have an idea where ammonia comes from and how it behaves in water and under different pH conditions. What we still need to find out is how fish deal with ammonia and what actually happens when ammonia is present in the water the fish live in. So, first a little bit more biology.

Basic fish biology – the gills

As mentioned above, the water solubility of ammonia is a big disadvantage for mammals and forces them to expend extra energy to get rid of the ammonia they generate; but it is a big advantage for fish because they do live in water and not only that, they ‘breathe’ it, too. 
The normal breathing, that is the uptake of oxygen and expulsion of CO2, happens in the gills. The gills are one of those many little engineering marvels in nature (see picture below) and generate a really large surface area where the blood can get very close to the water, just separated by a very thin membrane in the lamella. Here oxygen gets from the water into the blood and CO2 from the blood into the water. But it is also the main site where the fish gets rid of the ammonia that has been produced in the process of breaking down proteins (see previous sections). Because the concentration of ammonia in the blood is higher than in the water it diffuses from the blood into the water.

This is basically the same process that can be seen when you put a drop of ink into a glass of water: you have an area of high ink concentration and the surrounding water with no ink in it, yet after some time the whole glass of water will be colored by the ink. And it prefers to stay that way, you never see a glass of lightly colored water where suddenly all the ink decides to gather in just one area leaving the rest of the glass clear; in the same manner ammonia moves from the higher concentration in the blood out into the water and not the other way around.

And here we can see an immediate problem that arises when there is ammonia in the water – because the process of ammonia elimination relies on a difference in the ammonia concentration between the blood and the water, the ammonia in the fish blood suddenly has problems getting out. If there is a higher ammonia concentration in the water than in the blood the process is even reversed and ammonia is taken up into the blood. Not a good thing!

(The above image is from from www.BaileyBio.com – an excellent source if you want to learn more about basic biology)

Now if you know a little biology you may say: hang on, I have seen anatomical pictures of fish and there was a kidney in there somewhere, so what is that one doing then, if not generating pee?
We need to get back to our ink-in-water example for that one. The same we said about ink and ammonia wanting to distribute evenly in water is true for all other types of materials soluble in water as well, including salts. In fresh water fish the concentration of (all kinds of) salts inside the fish is actually higher than in the surrounding water. But as these salts are largely trapped inside the various cells of the body, they can’t easily get out. However, the water outside can relatively easily get in - so if the mountain won’t come to the prophet…. To avoid swelling up like a balloon and having all kinds of problems in the process, the fish need to somehow get rid of all the water coming in all the time. That is what the kidney does in a fish: it filters out all useful things from bodily fluids for recycling and excretes the excess water – so fish pee is actually pure water!

Molecular mechanisms of ammonia excretion

This section is going to get a little more technical, but I’ll try to explain what’s going on in a way that is hopefully easy to understand. You can skip this if it seems a bit too complicated.

Almost all the biochemical processes of breaking down proteins into those universal building blocks and energy units happen inside a cell. This means that all the waste ammonia needs to somehow get out of the cell and into the blood to get to the gills. Because under normal circumstances the blood of the fish is not directly exposed to water – they would have severe survival problems if it was – the ammonia needs to cross another layer of cells in the gills to get from the blood into the water passing through those gills. The basic mechanisms for getting out of a cell into the blood, from the blood into another cell, and from there into the water are the same, so I will only explain it in general terms and not for each individual hurdle the ammonia has to pass before it gets into the water.
The cells of all creatures are designed to keep things inside and nicely together but also to allow a level of control about what is getting into the cell and what gets out. For this purpose they are surrounded by a thin membrane made up out of various lipids (the biological equivalent of a plastic bag, really) studded with a number of little valves and pumps that move things in and out of a cell. 
In the picture below I have drawn the schematic representation of a cell and the cell membrane surrounding it. Let’s say the cell sits inside the gills and the outside is exposed to the water. The top case (A) illustrates the situation under normal circumstances where there is a higher concentration of total ammonia inside the cell and basically no ammonia in the water.

There are three different ways the ammonia can get from the cell into the water. The first and most simple is actually one that doesn’t really involve the cell itself. The cells making up the blood vessels (or other tissue in the body) are stacked quite tightly together so that the blood doesn’t leak out, but there are still tiny little gaps in between the cells and something as small as a single molecule of ammonia can slip through. Although it has to be said that the ammonia getting out this way makes up only a tiny proportion of all the ammonia moving from the gills into the water.

The second way is via the diffusion process described in a previous section. The only problem here is that there is a membrane in the way. As I pointed out the membrane is largely made up out of lipids, which are just different types of fat, and water doesn’t really mix well with fatty lipids. In chemistry terms the water is a ‘polar’ liquid, whereas the lipids are ‘non-polar’ and one just won’t go into the other. The ammonium ions, because they carry an actual charge, are even more polar than water, so they have no chance at all to get through that fatty membrane that stands between them and the water.
But here the ability of ammonia to exist partially as charged ammonium and as free ammonia comes in handy. As the free ammonia doesn’t carry any charge it is far less polar that the ammonium ion, in fact it is a lot less polar than water as well. This means that the ammonia can actually slip through the membrane and go back to being a happy ammonium on the other side of the membrane. As a charged ammonium it can’t get back into the membrane and because water is constantly passing through the gills the ammonium ion gets swept away before it can turn back to ammonium and eventually ends up outside the gills. The equilibrium between both ammonia and ammonium makes sure that there always is some free ammonia inside the cell that can move to the outside, driven by the difference in concentration between the outside and inside of the cell.

This process can be greatly helped by the right pH of the water passing by. The actual pH of fish blood and most tissues in the fish is about 7.2, so if the pH in the water is lower, this means that more of the ammonia will exists as ammonium, making it even harder for the ammonium to form ammonia and to get back into the cell.
Now, don’t go back to your tanks and adjust the pH to 6 in all of them. Please keep in mind that pH and ammonia excretion are only one aspect of fish health. Fish are adapted to their natural environment and fish from higher pH waters will run into all kinds of health troubles if the pH is too low. They may have less trouble getting rid of ammonia, but that is of no use to them when they are dead! 

The third possibility for ammonia to cross the cell membrane is the action of the aforementioned pumps and valves that are built into the membrane. Some of these have the ability to pick up ammonium ions on the inside of the cell and dump them on the outside. The advantage of this method is that it allows some control about what gets moved where; if more ammonium needs to be moved, the cell can make more of these pumps – within limits. The disadvantage is that these pumps require energy, which means that there is less energy available for other things that need to be done inside the cell, so the cell usually tries to get away with having as few of these pumps as possible.
The case B in the above figure illustrates the situation when there is a higher concentration of total ammonia in the water than inside the cell. With higher total ammonia now the equilibrium pushes things the other way. More total ammonium means more ammonia that can get through the membrane, but this time into the cell! The pumps will still work to a degree, but higher ammonium concentrations mean that the cell needs to spend a lot more energy pushing all that ammonium out again, but eventually won’t be able to keep up with all the ammonia sneaking in through the membrane. If the concentration of ammonium outside gets too high, the pumps will stop working altogether (called ‘substrate inhibition’). Ammonia now is free to go anywhere and the cell will start to die – and with it the fish, eventually.

Some general remarks about toxicity

When we talk about toxic effects on fish we will have to distinguish between two different cases. One is the acute toxicity, the fish dies or gets visibly sick within a short period of time after adding the toxin; the other is chronic toxicity, an effect that can only be observed after exposure to the toxin over an extended period of time, or effects that show up only very late (such as cancer). The effects themselves may be either transient, they go away again when the toxin gets removed, or permanent, i.e. the effects stay even after the toxin is removed.

Acute toxicity usually gets measured as LD50 which is the dose of the toxin at which 50% of the test subjects die (LD = lethal dose). When looking at LD50 data we also need to keep in mind which timeframe the study looks at; some studies may look at death after 1h while others may look at death after 24h so the LD50 values would be quite different. 

While acute toxicity is relatively easy to determine, chronic toxicity can be tricky and the studies need to be a lot more elaborate to get meaningful data. For example if I take a single fish and keep it in a tank with a fixed concentration of ammonia at a constant pH and temperature for 6 months, then do an autopsy on the fish and find that it had discolored gills, a tumor on the liver and a slightly bent spine. How can I be sure that the discolored gills were the effect of ammonia? How do I know that the fish did not have a genetic defect responsible for the spine? How do I know if the tumor was just a coincidence or not? How do I know that there is not something else in the water I use, e.g. heavy metals or pesticides that affected the fish? Was there always enough oxygen in the water? To exclude all these possibilities I will have to at least set up a control group that gets treated exactly the same just without the ammonia present, make sure all the water I use is really clean, that the equipment doesn’t pollute the water, and I will have to look at larger groups of fish for all experiments to make sure that I didn’t just pick one or two fish that were sick anyways. That makes the whole thing a lot more complicated – and expensive. An experiment on a single fish or a small number doesn’t really tell you anything much; it may be a good starting point to set up some proper experiments, but really only what is called ‘anecdotal evidence’, meaning someone reported they have seen something and may believe they know how it happened, but there is no conclusive proof.
One common way to measure chronic toxicity is to use an EC20 value, which is the concentration of the toxin at which 20% of the test subjects show some symptoms of poisoning after a specified time (EC = effective concentration).
Of course different species of fish will react differently to different concentrations of ammonia, so results from one species can’t be applied across all species.

Toxic effects of ammonia

Now – finally – let’s have a look at some the data that is out there describing the toxic effects of ammonia in fish. The symptoms of acute intoxication with high levels of ammonia are convulsions quickly followed by death. Acute toxicity of ammonia is largely due to its effect on the central nervous system. Funny enough, despite its importance for fish health the exact details of how ammonia affects the central nervous system are actually not really known. I guess no one is interested enough to find out – not that it is that easy, mind you.

The question of how much ammonia is too much is one of the important ones in commercial aquaculture and also in conservation when it comes to setting limits for what is advisable or allowed in waterways and/or commercial fish farms. Most published data is actually on commercially important species as well as tadpoles and some invertebrates such as daphnia, fresh water shrimp etc. that are used in ecology as indicators for water quality. The data on ornamental fish in the hobby is a lot scarcer (hobbyists don’t have the money to pay for expensive toxicology trials), so when discussing general toxic effects and ranges of tolerable or dangerous ammonia concentrations a lot of the data will be based on commercial species. 
A good compilation of data on ammonia toxicity can be found in the documents by the United States Environmental Protection Agency (USEPA) that are listed in the reference section and most of the data I will discuss comes from that source. Australia has largely adopted the conclusions and recommendations from the USEPA documents for our own regulations about allowable limits etc. 

Ammonia – acute toxicity

First, because it is much simpler to determine, some values for and discussion of acute ammonia toxicity. The table below shows the acute ammonia toxicity that has been determined for salmon. Because of its commercial importance and it is a wild species well known to the public – who hasn’t seen pictures of the migration of salmon up those rivers to get to their spawning grounds and provide food for countless bears in the process? – there is quite a bit of data available on salmon. It also is a species that tends to be more sensitive towards ammonia than other species, so it is a good species to look at because if we can get some ‘safe’ limits for salmon they are likely to work for other species as well.
The table below shows some experimentally determined LD50 doses for salmon at 30°C at different pH values and two different exposure times; it shows the concentration of total ammonia (NH3 + NH4+) and the calculated amount of free ammonia (NH3) (for the given pH at 30°C, calculated using the equation in the chemistry section) at which 50% of the salmon die, either after 1 hour or after 4 days.

There are a few interesting things in this table. The first is the huge difference in total ammonia lethal to fish if we compare exposure for 1 hour and 4 days. If we only looked at say the value for acute toxicity of total ammonia after 1h at pH 6.5, it is a relatively large number – 14.3ppm. But then at the same pH about 1/20th of that dose (0.73ppm) is still enough to kill the fish over 4 days. 
The second is that the amount of total ammonia required to kill a fish gets lower and lower the higher the pH is, which goes nicely with the calculations that we have done before demonstrating that there is more free ammonia present at higher pH. But if we look at the details, the amount of total ammonia for the 4 day LD50 is the same for pH 6.5, 7.0, and 7.5 and starts to drop only after that. That seems a bit odd given that we know that there is more free ammonia present at higher pH. If we look at the calculated values of free ammonia that are in the column on the right, we can see that e.g. for the 1h LD50 at pH 6.5 there is 0.036ppm free ammonia and at pH 8.5 there is 0.26ppm, more than 7 times as much to get the same effect (50% of the fish die)! It almost looks like the fish can actually tolerate more free ammonia at higher pH than they can at lower pH. How can that be?? 

Well, it shows that things are not quite as simple as they are often made out to be: ‘ammonia problem? – just lower the pH, all ammonia will be ammonium and all will be A-OK’. Not quite. This data clearly shows that it is not just the free ammonia (NH3) that has a toxic effect; ammonium (NH4+) is toxic to fish as well!! 
The reason is that under certain circumstances ammonium can actually get through the cell membrane in the gills and into the fish. There are two ways how that can happen. One is as described in the figure further above where a high concentration of ammonium outside the cell leads to ammonia getting ‘pushed’ in; as soon as some ammonia is in, the chemical equilibrium requires that some of the ammonium on the outside gets converted to ammonia, which can then be pushed into the cell etc. And, as already pointed out, high ammonium also makes it difficult for the ammonia from inside the fish to get out. The other reason ammonium can get in is that an ammonium ion is about the same size as a potassium ion (K+) and some specialized pumps that bring in the potassium that the fish needs accidentally also bring in ammonium. 
In fact, once inside the fish it is actually the ammonium ion (NH4+), not the free ammonia (NH3) that is the cause for the neurotoxic effects – ironic, isn’t it.

Events that can lead to levels of ammonia in natural waterways high enough to lead to acute toxic effects are often man-made. Of course there always will be stagnant waters where decaying organic matter can lead to high levels of total ammonia, or the herd of 25,000 wildebeest stopping at a small river for a drink and a pee. But in many cases high ammonia is caused by runoff from agriculture and farming. Fertilizers and poo are packed with nitrogen, which makes them good for plants, not so good in waterways. And of course there is intense aquaculture where tens of thousands of fish are kept in close proximity, producing a lot of waste that gets flushed away into the surrounding waters.

The USEPA has come up with a set of guidelines that they called ‘The National Criterion for Ammonia in Fresh Water’, it defines a ‘Criterion Maximum Concentration’ as the ‘one-hour average of total ammonia nitrogen (in mg N/L) does not exceed more than once every three years on average, the CMC calculated using the following equation….’ The exact equation is not really important here, and the mg N/L is almost the same as ppm total ammonia, so I will continue to use that to avoid confusion. 
There is a table in the USEPA document that details the CMC for different pH values; e.g. the CMC for pH 7.5 is about 13ppm for waterways with salmons present and 19ppm without. That is a bit higher than the values from the previous table. Apart from the fact that different studies can lead to slightly different results, the acceptable CMC basically says that it is ok to have total ammonia values high enough to kill half of all salmons and a proportion of all other aquatic life as long as it doesn’t happen more often than once every three years…. 

Ammonia – chronic toxicity

Whereas things are relatively clear-cut with acute toxicity (the fish either die or they don’t) it gets a lot murkier when it comes to chronic toxicity. We can’t just go and ask a fish: ‘So, Mr. Salmon, how are you feeling after 30 days of 0.02ppm total ammonia? Any particular problems? Breathing still ok, no pain in the gills?’ 
To be able to determine chronic effects there need to be clearly observable changes in the fish. Some of these may be easy to detect, such as some changes to the skin, others may require a detailed histopathological analysis – the fish will need to be cut up and all organs carefully inspected under a microscope, down to a cellular level if you want to be really exact about it. Because that is quite a lot of work there is not really that much detailed data around, so I’ll stick to just one or two of the better examples to discuss some of the data available. If you want to know more look at the articles in the reference section.

If we look at some of the official documents that try to define maximum values for ammonia in waterways with regards to chronic effects it starts to get a bit complicated. I’ll quote from the 1998 USEPA document that also forms the basis for Australian legislation. To come up with some form of standard formula that can be applied to various waterways and across different species, data from the determination of different EC20 values (= concentration at which 20% of fish show a measurable effect) was used to calculate what they called a ‘Criterion Continuous Concentration’ (CCC). According to that definition the CCC is (in bureaucracy speak) means that the ‘thirty-day average concentration of total ammonia nitrogen (in mg N/L) does not exceed, more than once every three years, the CCC calculated using the following formula……and the highest four day average within the 30-day period does not exceed twice the CCC.’
That does sound not too bad, maybe? Apart from the fact that this basically requires permanent monitoring. How would you be able to tell otherwise? I am not sure how easy that would be to achieve that for all waterways that may be affected by e.g. run-off from fertilized fields.

So what are the actual values for this CCC? There is a table in the USEPA document that details the CCC for different pH values; the CCC for pH 7.5 is about 2.3 ppm for example. Wait a minute! Didn’t we just see in the part about acute toxicity that the dose of total ammonia that kills 50% of all salmon in 1 hour at pH 7.5 is 7.3ppm and that even 0.74ppm total ammonia over 4 days kills 50% of salmon? How can that CCC be 2.3?? Well, two things come into play. 

One is the fact that the CCC is derived as an average over all species (fish, invertebrate, etc.) and many of them are less susceptible than salmon so the average will be higher than for salmon alone. So the CCC does not look at the most sensitive species but at an average over a number of different species. If you happen to be a salmon, well, that’s bad luck then. 
The second one is the definition of the EC20.One would expect that to set levels for chronic toxicity you’d look at things like signs of stress or gill damage, right? Well, the ‘toxic effect’ the EPA has been looking at in the relevant studies to determine EC20 and calculate the CCC is – death. That is right, the only difference between looking at values acute and chronic toxicities is that in one case 50% of the fish die, in the other case ‘just’ 20%. That is like looking at long-term toxic effects of lead in water and say ‘it is safe to ingest this much lead because less than 20% of people die from it over the course of a few days’!! I bet all the lobbyists for the big agriculture and other industries have been doing their victory dances and had a good laugh after this came out!

Well, it seems obvious, to me at least, that those CCC values are a bit dodgy. So, instead let’s have look at what other data is out there when it comes to chronic ammonia toxicity in fish.
There is actually some evidence that total ammonia concentration as low as 0.001-0.006ppm can have a negative effect on aquatic life – this is below the detection level of your normal aquarium test kit! There is quite clear evidence that with increased ammonia concentrations you will get decreased survival, growth, and reproduction. When these effects kick in exactly will depend on the individual species and on environmental factors like pH, temperature, stress levels, to name just a few. Funny enough those studies are actually discussed in those same EPA documents.
For some salmon species inhibition of growth of fry was shown to occur at total ammonia concentrations as low as 0.002-0.006ppm at pH levels between 6.4 and 7.1. For some non-salmonid species those levels can be quite a bit higher, e.g. 0.11ppm for bluegill (Lepomis macrochirus) at pH 7.8 and 0.32ppm for channel catfish (Ictalurus punctatus) at pH 7.9.

Some more detailed studies are available for rainbow trout. For example, when exposed to 0.02ppm total ammonia for 2 weeks rainbow trout fry showed changes in the skin and at 0.06ppm total ammonia freshly hatched fry showed damaged kidneys, and 50% of fish exposed to 0.06ppm at pH 7.4 for 72 days died – and those are ammonia still below what most test kits can detect!

One very interesting study looked at the effect of low ammonia concentrations on rainbow trout over several generations: they were all exposed to total ammonia concentrations ranging from 0.008-0.06ppm, the parents for 11 months, the first generation (F1) for 48months and the second generation (F2) for 5 months. Fish from all generations were carefully analyzed at different time points looking at tissue samples from spleen, heart, gills, liver, and kidney. 
Although there were no visible effects on growth and reproduction in general, there were clear effects observed for gills and kidneys that got worse the higher the concentration of total ammonia was, starting from 0.02ppm. The most profound effects were in the gills, which makes sense as the gills have a huge surface area and are in constant contact with the water. The effects were similar to people suffering from emphysema and affected swimming performance and stamina of the fish. The effects were similar for all three generations but the F2 generation also suffered from a protozoan infection. It was, however, not clear if this was due to an increased susceptibility because of the exposure to ammonia or some other reason. But that is probably just scientists being quite strict in the interpretation of data and careful in what they are saying; it does seem likely that there is a connection.

It is not all bad news, though. Some fish of the F1 generation were moved to clean water after 7 months at 0.05ppm total ammonia showing the above signs of gill damage and they completely recovered over the course of a few months. So it looks like the damage caused by exposure to low level ammonia concentrations is reversible when fish are returned to clean water.
We also need to keep in mind that most of the discussion above was for salmon and trout species, which are more sensitive to ammonia than many other fresh water species. But having said that, even if we assume the average fish to be 10x less sensitive to ammonia the levels of total ammonia tolerable to fish are actually not that high and for chronic toxicity damaging levels of ammonia can still be below the detection limits of most standard ammonia test kits in the hobby.

Ammonia testing and ammonia removers

I won’t go too much into detail here or discuss advantages and disadvantages of different brands; there is plenty of information available and everyone has their own favorites and for the most part there is not too much difference between various brands.

For the determination of ammonia levels there are three different types of ammonia test kits around, two for the determination of total ammonia (NH3 + NH4+) and one for free ammonia (NH3). The two types of test kits for total ammonia use different chemical reactions to detect ammonia. 
The older one uses Nessler’s Reagent which gives a yellow to orange color that is compared to a color chart to give a reading for different ammonia concentrations. Although quite accurate under the right circumstances it is not so widely available anymore as it does contain mercury salts that are quite toxic and shouldn’t really go down the drain. The disadvantage with this type of test kit is that it is affected by some dechlorinators and additives to remove ammonia (such as Prime) and gives a false positive reading.

The other type of test kit for total ammonia uses a salicylate reagent that gives yellow to green colors that can be compared to a color chart to give a reading of total ammonia present. The advantage of the salicylate test is that it does not contain toxic mercury salts and is also not affected by dechlorinators or additives to neutralize ammonia. That means that a salicylate test will still show a reading for total ammonia even after adding Prime, for example, as the salicylate reagent liberates the ammonia bound by Prime and makes it available for testing.
Using the formula given in the section about ammonia and pH the amount of free ammonia can be calculated from the values for total ammonia, pH, and temperature.

The third type of ammonia test is specific for free ammonia. It usually comes as a little disk to attach inside the tank and the colors on the disk will give a reading for free ammonia present. This is useful as it gives a direct reading for the most toxic form of ammonia without having to calculate anything. However, as it does not respond to ammonium ions at all, for low pH tanks this can lead to a situation where there is a high (and toxic) ammonium concentration present which the test does not show.

There are number different ammonia removers around. The most common ones are in a liquid form and contain an aldehyde, such as formaldehyde or glutar aldehyde, which reacts with the free ammonia to form what is called an imine (sometimes also called iminium). These imines are not directly toxic to fish so these ammonia removers are quite good to keep ammonia under control in an emergency or while the tank has not stabilized. Because of some toxicity issues with formaldehyde (and other aldehydes as well) it is usually not the free aldehyde that is used but the bisulfate complex or the dimethylacetal. Although this leads to fast neutralization of ammonia and greatly reduces toxicity to fish, the resulting imines should still be removed by successive water changes. The reason for this is the fact that the imines are not that stable in water and slowly break down releasing some of the ammonia. While this is great when cycling a tank as it still allows the nitrifying bacteria to develop, we have seen in the previous section about ammonia toxicity that even very small amount of ammonia can have long-term detrimental effects on fish health. 

Another form of ammonia remover is a solid that can be placed in the filter or a spot in the tank with strong water flow. These are usually zeolites, a type of mineral that has the ability to ‘capture’ ammonia from the water, or ion exchangers, a synthetic material that does a similar job. Both zeolites and ion exchangers actually capture only ammonium (NH4+) and their efficiency is dependent on pH and can also be affected by salt in the water, this is especially true for ion exchangers. Hence this method of ammonia removal is generally only suitable for freshwater tanks. In addition to that their capacity to remove ammonia is limited, so they eventually stop working. But again, they are good when dealing with an acute problem, although they neutralize/remove ammonia not quite as fast as the liquid ammonia removers. 

So, that’s it in a (large) nutshell: all that you ever wanted to know about ammonia and probably a bit more….

References

USEPA, Draft 2009 update aquatic life ambient water quality criteria for ammonia – fresh-water, Washington, D.C.
(http://water.epa.gov/scitech/swguida...2009update.pdf)
USEPA, 1998 Update of ambient water quality criteria for ammonia, Washington, D.C. 
(http://water.epa.gov/scitech/swguida...ia_nh3_rpt.pdf)
USEPA, Ambient Water Quality Criteria for Ammonia-1984, Washington, D.C.
(http://water.epa.gov/scitech/swguida...mmonia1984.pdf)
Ammonia toxicity in fish, D.J. Randall and T.K.N. Sui, Marine Pollution Bulletin, 45(2002)17-23 (http://dx.doi.org/10.1016/S0025-326X(02)00227-8)
Gas transport and gill function in water-breathing fish, S.F. Perry, A. Esbaugh, M. Braun, K.M. Gilmour, Cardio-respiratory control in vertebrates (M.L. Glass and S.C. Woods, Eds.), 2009, pp. 5-42
The multifunctional fish gill: dominant site of gas exchange, osmoregulation, acid-base regulation, and excretion of nitrogenous waste, D.H. Evans, P.M. Piermarini, K.P. Choe, Physiology Reviews, 85(2005)97-177 (http://0-physrev.physiology.org.libr...t/85/1/97.full)
Ammonia excretion and urea handling by fish gills: present understanding and future research challenges, M.P. Wilkie, Journal of Experimental Zoology 293(2002)284-301 (http://www.wilfridlaurier.ca/documen...3,_284-301.pdf)

A detailed description for the situation in a marine environment:
USEPA, Ambient water quality criteria for ammonia (saltwater)-1989, Washington, D.C.
(http://water.epa.gov/scitech/swguida...iasalt1989.pdf)

There are plenty of websites out there with good information – but also some with not so good information (aka rubbish), so be critical; the following I have used in the preparation of this article

http://water.epa.gov/scitech/swguidance/standards/criteria/aqlife/ammonia/technical.cfm
http://www.environment.gov.au/resource/australian-and-new-zealand-guidelines-fresh-and-marine-water-quality-volume-1-guidelines
http://www.daff.qld.gov.au/home.htm
http://www.ca.uky.edu/wkrec/ManagingAmmonia.pdf
http://dataguru.org/misc/aquarium/AmmoniaTox.html
http://www.thekrib.com/Chemistry/ammonia-toxicity.html
http://edis.ifas.ufl.edu/fa031
http://www.skepticalaquarist.com/

This article was originally published in the AquariumLife forum thread: Ammonia in the freshwater aquarium started by regani View original post