A simple continuous and fail-safe water-change system

ian_m

Global Moderator
Staff member
Joined
25 Jan 2012
Messages
5,221
Location
Eastleigh
I’ll prepare some standard solutions and check the accuracy of the API test
Well that can be one issue with test kits, repeatability with known strength solutions, in that different manufacturers kits give different readings for the same solution.

The main problem with hobby test kits is ionic interference. Other salts in the tank water can cause test kit false readings. Chloride ions can influence nitrate results, chlorine present in untreated tap water causes a lot of test kits to under read and dechlorinator can cause test kits to over read.

If you insist on wasting time and money testing your water using test kits, then proper "big boy" kits are the way to go. These are generally not affected by other ions.

Nitrate
https://uk.hach.com/nitrate-test-kit-model-ni-11/product?id=26427780260&callback=pf
Phosphate
https://uk.hach.com/test-kit-phosph...del-po-19a/product?id=25114225337&callback=pf
Chlorine
https://uk.hach.com/chlorine-free-c...66f-pk-100/product?id=25114235552&callback=pf
Anyway you get the idea.
 

dw1305

Expert
Joined
7 Apr 2008
Messages
11,277
Location
nr Bath
Hi all,
The main problem with hobby test kits is ionic interference.
That is the main problem.

Strangely it is easier with sea water, because you can take into account a <"known amount of chloride ions (Cl-)">, about 19 ppt (1.9% by volume), but in freshwater you have different amounts of anions, dependent upon whether the water is hard or soft etc.

There are a number of issues particularly with the nitrate ion (NO3-), and one of these is that <"nearly all nitrate compounds are soluble">, and if we want to use a colorimetric test (one that uses change of colour), we need the end product (the thing we measure) to be a coloured compound.

Colour kits work by nitrate reduction to nitrite (NO2-), which forms a number of coloured compounds. The two normal methods are <"cadmium reduction" and "azo dye" formation">. Cadmium reduction has the <"cadmium (Cd) issue">, so the "azo dye method" is usually used (a red colour shows the concentration of NO3, with cadmium reduction you get an amber colour reading). If you have an amber coloured kit, you really need to shake the bottle hard to make sure you get the full colour development.

You tend to get a slightly more accurate results with an ion selective electrode (ISE), but even then you can get problems with ion interference (the ion selective bit is a semi-permeable membrane which filters out ions which are the "wrong" size). Ion selective electrodes are quite expensive bits of kit.

This was a major reason for using <"conductivity and the Duckweed Index"> as a proxy for nutrient status, it isn't perfect but it probably gives you a better idea across a whole range of water conditions.

cheers Darrel
 

David Cottle

Member
Thread starter
Joined
9 Jan 2018
Messages
36
Location
NIagara-on-the-Lake ON Canada
If you insist on wasting time and money testing your water using test kits, then proper "big boy" kits are the way to go.

That's a pretty broad statement o_O

When I did a fishless cycling of my new tank, the API ammonia test kit showed concentrations consistent with the amounts of aqueous ammonia I had added to the tank. Subsequent nitrite measurements were consistent with the stoichiometry. Threadfin Rainbowfish added after the test showed zero nitrites did well. The tests apparently served their purpose.

As for the nitrate test accuracy, I appreciate the alarms raised here. I would like an idea of actual nitrate concentrations, and might get a Hach kit if I find that API testing of samples of my (C-filtered) water with measured amounts of my fertilizer solution gives inconsistent results. Don't need absolute accuracy - approximate should do. The Hach test, at $1+ per test, might be overkill for routine monitoring.

Colour kits work by nitrate reduction to nitrite (NO2-), which forms a number of coloured compounds. The two normal methods are <"cadmium reduction" and "azo dye" formation">. Cadmium reduction has the <"cadmium (Cd) issue">, so the "azo dye method" is usually used (a red colour shows the concentration of NO3, with cadmium reduction you get an amber colour reading). If you have an amber coloured kit, you really need to shake the bottle hard to make sure you get the full colour development.
IMG_0840.JPG


The pic shows an older API nitrate test bottle #2 that I cut open to have a look inside. I assume the silver-grey residue on the bottom is metallic Cd powder. It's loose - not caked. I taped the bottle closed, shook it vigorously, and found most of the powder in suspension when I reopened the bottle.

A bit of reading of "API nitrate test reviews" indicates that test results are technique-sensitive, which is probably the case with all hobby tests, and perhaps more-so with this one. I suspect that shaking of the final sample-reagent mixture for the time specified is probably more crucial than prolonged shaking of reagent bottle #2 before dispensing - as long as a surplus of Cd is dispensed for the test. I calculate the Cd equivalent of 40 ppm nitrate, e.g., in a 5-ml sample to be 0.36 mg.

If I'm going to check the accuracy of the test with my water, I might as well check results with varied shaking to calibrate my technique.

I'll also get a conductivity meter for additional information and will probably use it eventually for monitoring. At this point, however, I'm not sure that conductivity alone will tell me how much nitrate I have, versus an accumulation of other fertilizer components such as sulphates.

And I'll read up on your Duckweed Index :)

Thx.
 
Last edited:
Joined
26 Feb 2013
Messages
3,379
I'll also get a conductivity meter for additional information and will probably use it eventually for monitoring. At this point, however, I'm not sure that conductivity alone will tell me how much nitrate I have, versus an accumulation of other fertilizer components such as sulphates.

Why do you need to know the level of nitrates? If the conductivity is increasing from the value of your tank water, and you are not adding artificial fertiliser,then water quality is deteriorating. If the nitrate test at the same time has shown low levels of nitrates, and you decrease the water changes amount/frequency because of that, then you're on the wrong path.

The nitrate test will be inconsistent with the TDS increase, depending on the tank set up, especially in a planted tank. The last time I did a nitrate test was somewhere around 2014. The nitrate in the tank was barely 5ppm. That's because I had a heavily planted tank at the time in which I didn't dose fertiliser. But I promise you TDS was rising steadily if I don't water change, regardless of levels of nitrates. I once stopped doing water changes for about 5 months, unhappy fish, TDS doubled over that period!
 

dw1305

Expert
Joined
7 Apr 2008
Messages
11,277
Location
nr Bath
Hi all,
The pic shows an older API nitrate test bottle #2 that I cut open to have a look inside. I assume the silver-grey residue on the bottom is metallic Cd powder. It's loose - not caked. I taped the bottle closed, shook it vigorously, and found most of the powder in suspension when I reopened the bottle.

A bit of reading of "API nitrate test reviews" indicates that test results are technique-sensitive, which is probably the case with all hobby tests, and perhaps more-so with this one. I suspect that shaking of the final sample-reagent mixture for the time specified is probably more crucial than shaking of reagent bottle #2 before dispensing - as long as a surplus of Cd is dispensed for the test. I calculate the Cd equivalent of 40 ppm nitrate, e.g., in a 5-ml sample to be 0.36 mg.

If I'm going to check the accuracy of the test with my water, I might as well check results with varied shaking to calibrate my technique.
I don't which bottle is which in the API test, but that seems likely, and a good shake should work.
I would like an idea of actual nitrate concentrations, and might get a Hach kit if I find that API testing of samples of my (C-filtered) water with measured amounts of my fertilizer solution gives inconsistent results. Don't need absolute accuracy - approximate should do. The Hach test, at $1+ per test, might be overkill for routine monitoring.
I would like to know as well.

We are lucky in the UK that we can get accurate NO3 readings from our water supplier, but after that it becomes more problematic. @alto is a scientist and he doesn't see any reason why you can't get accurate results with kits, if you follow the <"proper protocol and the scientific method">.

I'm going to stick with using plant health and colour, it isn't perfect, but it has a <"basis in science">.

cheers Darrel
 

David Cottle

Member
Thread starter
Joined
9 Jan 2018
Messages
36
Location
NIagara-on-the-Lake ON Canada
Why do you need to know the level of nitrates?

Well, for several reasons. I'm new at this, as well as curious and somewhat analytical by nature. I add fertilizer daily, and am trying to stay within a nitrate range of 5-20 ppm that is good for plant growth but not stressful for fish. If my measurement is at the high end of this range, I'll skip a daily dose; if it's low, I increase the dosage.

With continuous water change and an increase in conductivity, I wouldn't know without other information whether to increase the purge rate or reduce the fertilizer dosage :nailbiting:. With enough information and more experience, I'll have a better feel - like you - for how best to regulate it :)
 

Edvet

Global Moderator
Staff member
Joined
15 Aug 2013
Messages
5,149
Location
Lelystad, Netherlands
trying to stay within a nitrate range of 5-20 ppm that is good for plant growth but not stressful for fish
There are plenty scientific papers where high nitrate levels ( up to 500 ppm) are not detrimental to fish health. However if these high levels are caused by the nitrogen cycle, the higher nitrite and ammonia levels are dangerous. Adding NO3 isn't.
 
Joined
26 Feb 2013
Messages
3,379
I agree with Edvert. One forgets that in order for a tank to build up nitrates, there must have been sufficient ammonia produced at the same time. And because people test ammonia as zero in a cycled tank, they tend to forget ammonia is actually constantly being produced/converted to less harmful compounds 24/7 in a fish tank and it may spike, whether you happened to test at that exact moment or not. Go blame high nitrates for your fish problems when in the first place there is high levels of ammonia being produced...whether detected or not......

That process is heavily oxygen demanding(oxygen is a limitation in fish tanks) and the by-products of aerobic and anaerobic nitrification are numerous and mostly harmful. Nitrate in a planted tank is actually never a problem. Plants can also use ammonia directly as nitrogen source, which is one of the great benefits of plants, thus reducing heavy nitrification.
 

dw1305

Expert
Joined
7 Apr 2008
Messages
11,277
Location
nr Bath
Hi all,
There are plenty scientific papers where high nitrate levels ( up to 500 ppm) are not detrimental to fish health. However if these high levels are caused by the nitrogen cycle, the higher nitrite and ammonia levels are dangerous.
It is like @Edvet says this is quite an interesting area, because it is only really aquascapers who add NO3 to water containing fish.

In aquaponics (or organically polluted water etc.) the NO3 is there as the "smoking gun" of earlier elevated levels of ammonia (NH3) and nitrite (NO2), and usually goes hand in hand with <"elevated levels of orthophosphate (PO4---)">.

Unfortunately we don't know at what level nitrates become a problem for most ornamental fish, although we could assume that it would be higher for most of these than it would be for <"aquacultured salmonid fish">. There are some more references in <"What is nutrient toxicity"> and <"Nitrates">
I add fertilizer daily, and am trying to stay within a nitrate range of 5-20 ppm that is good for plant growth but not stressful for fish. If my measurement is at the high end of this range, I'll skip a daily dose; if it's low, I increase the dosage.
This is back to why I like plants as a visual indicator of nutrient content, you have a negative feedback loop where increased fixed nitrogen leads to increased plant growth, which leads to lower fixed nitrogen levels. All nutrients can be limiting (this is back to <"Liebig's law of the minimum">) but a floating plant takes carbon (CO2) out of the equation and then plants need about an order of magnitude more nitrogen (N) and potassium (K) than phosphorus and about three times as much phosphorus (P) as magnesium (Mg) etc., again there is a more detailed discussion in the <"Nitrate"> thread.

Personally I like (and have) low nitrate levels. How do I know this if I don't test? Because I have floating plants that don't ever grow very quickly, or ever look brilliant green, and I have low conductivity tank water. If you only have 100 microS of conductivity, you have very few ions of any description.

cheers Darrel
 
Joined
26 Feb 2013
Messages
3,379
Personally I like (and have) low nitrate levels. How do I know this if I don't test? Because I have floating plants that don't ever grow very quickly, or ever look brilliant green, and I have low conductivity tank water. If you only have 100 microS of conductivity, you have very few ions of any description.

I sort of "measure" nitrogen the same way, by the condition of the plants, but for me floaters completely die off, each time I tried keeping them. So I could not use the "Duckweed method" :) Then again I've also been keeping emersed plants for years, which seem to out-compete floaters for nutrients. Either way, my tanks are constantly nitrogen deficient. There's a couple of pictures I just snapped...



 

dw1305

Expert
Joined
7 Apr 2008
Messages
11,277
Location
nr Bath
Hi all,
I assume high levels make more algae, even if they don't bother the fish?
A bit of an unknown really, advocates of EI will tell you most forcibly that it doesn't. In the natural environment high levels of phosphorus and nitrate are definitely the markers of eutrophication, and often occur together with an algal bloom.

Physiologically the green algae belong to the same clade (<"Viridiplantae or Chlorobionta">) as all the higher plants and have the same photosystems and basic physiology. Because of this I can't see why conditions that favour the growth of the plants you want, doesn't also favour the growth of the plants you don't want ("algae").

Personally I don't mind a bit of algae, in fact I'd actually like a bit more <"periphyton or aufwuchs"> in the tanks.

cheers Darrel
 

David Cottle

Member
Thread starter
Joined
9 Jan 2018
Messages
36
Location
NIagara-on-the-Lake ON Canada
Physiologically the green algae belong to the same clade (<"Viridiplantae or Chlorobionta">) as all the higher plants and have the same photosystems and basic physiology. Because of this I can't see why conditions that favour the growth of the plants you want, doesn't also favour the growth of the plants you don't want ("algae").
I've read that heavy planting reduces algae in a planted tank. A common explanation is that the plants out-compete the algae for nutrients - which makes no sense to me whatsoever.

A more plausible explanation is that heavy planting creates conditions that are unfavourable for algae growth, such as postulated here: https://www.ukaps.org/forum/threads/what-exactly-causes-bba-part-2-bacterial-imbalance.38375/
(haven't figured out how to create a link yet :()
 
Last edited:
Joined
26 Feb 2013
Messages
3,379
I've read that heavy planting reduces algae in a planted tank. A common explanation is that the plants out-compete the algae for nutrients - which makes no sense to me whatsoever.

I have personally seen my iron deficient green spot algae come to life once I dosed iron. It had gotten as pale as the plants I was trying to fix. I've no idea who out-competed who for iron :lol:
 

dw1305

Expert
Joined
7 Apr 2008
Messages
11,277
Location
nr Bath
Hi all,
I have personally seen my iron deficient green spot algae come to life once I dosed iron.
The algae can show a rapid response to iron addition, if iron is deficient, because it has no internal plumbing and Fe++(+) ions will diffuse into each cell.

Higher plants won't show that instant response, because they can't transport the iron to already grown tissue, it can only effect the new leaves developed after iron was non-limiting.
I've read that heavy planting reduces algae in a planted tank. A common explanation is that the plants out-compete the algae for nutrients - which makes no sense to me whatsoever. A more plausible explanation is that heavy planting creates conditions that are unfavourable for algae growth,
The real answer is we don't know how heavy planting works, but it definitely does reduce algae in low tech tanks.

cheers Darrel
 

zozo

Member
Joined
16 Apr 2015
Messages
7,545
Location
Netherlands
A common explanation is that the plants out-compete the algae for nutrients - which makes no sense to me whatsoever.

That's also a freedom of (mis)interpretation.. :) if the same sentence was put more to the point with "That enough healthy growing plants out-compete algae for nutrients"Would already make a lot more sense, it's a logical sum. Still doesn't say that if you have enough healthy growing plants you never will grow algae. :)
 

David Cottle

Member
Thread starter
Joined
9 Jan 2018
Messages
36
Location
NIagara-on-the-Lake ON Canada
conductivity meters have the advantage of giving consistent results over a large range of values, you get a linear response from RO water ~4 microS all the way up to sea-water at 53,000 microS. You really can just dip the meter in a take a reading, you want a low range meter, "Hanna" is a good make.
Darrel, is 0-1999 microS/cm a suitable range? And how often do you calibrate these?
 

dw1305

Expert
Joined
7 Apr 2008
Messages
11,277
Location
nr Bath
Hi all,
Darrel, is 0-1999 microS/cm a suitable range? And how often do you calibrate these?
Yes that is fine, you won't get above 1999 microS.

They differ from pH meters in that they only need re-calibration fairly infrequently (every three months or so) and you can make up the <"calibration standard your-self">, from potassium chloride (KCl) 0.01 M = 1411 µS at 20oC. I make up a 0.1M KCl solution and then dilute it (100 cm3 0.1M KCl made up to 1000cm3 with DI water)

When you want to take a reading you turn the meter on, dip it in the tank water and give it a swirl and take the reading.

cheers Darrel
 

David Cottle

Member
Thread starter
Joined
9 Jan 2018
Messages
36
Location
NIagara-on-the-Lake ON Canada
They differ from pH meters in that they only need re-calibration fairly infrequently (every three months or so) and you can make up the <"calibration standard your-self">, from potassium chloride (KCl) 0.01 M = 1411 µS at 20oC. I make up a 0.1M KCl solution and then dilute it (100 cm3 0.1M KCl made up to 1000cm3 with DI water)
Great!

I have KNO3 (and some other salts from Green Leaf Aquariums), not KCl. I can probably get KCl at the pharmacy.

Thanks :)
 
Last edited:

dw1305

Expert
Joined
7 Apr 2008
Messages
11,277
Location
nr Bath
Hi all,
I have KNO3 (and some other salts from Green Leaf Aquariums), not KCl. I can probably get KCl at the pharmacy.Thanks :)
They sell KCl as "sodium free salt" in the UK, but it isn't always pure KCl (it often has MgCO2 as dessicant), and may even contain some NaCl.

I'm not sure if these sell to Canada (it is a USA based company, and I don't know anything about them), but this sort of thing <"Now Foods - potassium chloride powder">.

If you can get pure KCl it is 0.746g KCl in one litre to give 1411 microS. I keep the 0.1M KCl stock (7.46g in 1000cm3) in a stoppered bottle, but I only use the 1:10 diluted calibration solution once, and I use the whole large volume for calibration. The reason for this is that they aren't like pH buffers, any drips of lower conductivity water will dilute the standard and change the microS reading.

I found this out the hard way when I bought some combined pH/Conductivity meters for field work. The same probe measured both pH and conductivity and you store the electrode in 4M KCl. Even the smallest trace of this storage solution left on the probe meant that you couldn't calibrate the meter for conductivity, and these particular units required calibration every time they were turned on, because they were pH probes.

If you can get pure NaCl? You need 0.491g (491mg) L-1 NaCl to give a 1000 µS standard.

cheers Darrel
 
Last edited:
Top