One of the most important issues facing marine aquarists is providing a suitable environment for their aquaria's inhabitants. Among the important properties for a marine environment's suitability is the water's salinity level. Water that is either too saline or not saline enough can be stressful or lethal to many organisms. Deciding what salinity level to maintain in a reef aquarium can be a complicated task, especially if the organisms come from different environments. This article includes a brief discussion of how to select an appropriate salinity target, but selection is not the main purpose of this article.

Monitoring the salinity level is an important issue itself, wholly apart from deciding what salinity level to target in an aquarium. Fortunately, a number of different methods for monitoring salinity are available to aquarists, including specific gravity (via hydrometers), refractive index (via refractometers), and conductivity (via electronic meters). In order to get the most out of any one of these methods, however, aquarists must have confidence that it provides reliable information. Each method can provide perfectly adequate information for aquarists, assuming that the device used is properly manufactured, calibrated, and used. Alternatively, each device can provide misleading information if any of these factors is not optimal.

Calibration of an analytical instrument is the best way of ensuring that the information that it is providing is accurate. While there are many ways to calibrate instruments for salinity determination, the simplest is to test the instrument using a solution with a known salinity. In that case, it is best if that standard has a salinity close to the samples likely to be tested, which in this case is close to natural seawater. If the instrument reads the appropriate value, it will be suitable for use by the aquarist. If not, then the device, or the aquarist's interpretation of the results, might need to be altered ("calibrated") to give the correct reading.

Aquarists can purchase commercial calibration standards that will permit calibration of each of these methods to a high degree of accuracy. These standards, however, can be expensive, and in many cases, complicated to use. One reason that commercial standards can be difficult to use, for example, is that they usually come standardized to units of measurement related to the technique, rather than to the unit of measurement the aquarist needs: conductivity for conductivity meters, refractive index for refractometers, and density or specific gravity for hydrometers.

Unfortunately, the reef aquarium hobby has seen a rapid expansion of devices for measuring salinity that report in units that they are not actually measuring. For example, how can a hobbyist use a refractive index standard (such as a solution with a known refractive index of 1.3850) when the refractometer reads in units of specific gravity?

This article describes a series of homemade calibration standards that can be made from sodium chloride (table salt) and purified freshwater. These solutions can be used to calibrate refractometers, hydrometers, and conductivity probes. While not likely to be as accurate as commercial standards (depending on the ability to accurately measure weights and volumes), they will be adequate for most reef aquarium purposes. At the very least, they can be used to prevent a seriously defective device from causing an aquarist to provide a grossly inappropriate salinity level in a reef aquarium.

Table 1 shows the relevant properties of seawater as a function of salinity. In order to make a standard for each method, it is necessary to determine what concentration of sodium chloride solution matches the appropriate property of seawater.^1-3 In this article, I will use solutions with the same properties as seawater with a salinity of 35 PSU (often written as S=35). PSU is an acronym for practical salinity units, which is essentially a modern replacement for ppt, since salinity is no longer defined as directly relating to solids in the water, but rather by its conductivity. How each standard is made and used is detailed for each of the different methods in subsequent sections.

General Salinity Discussion

As far as I know, there is little evidence that keeping a coral reef aquarium at anything other than a natural salinity level is preferable. It appears to be common practice to keep marine fish, and in many cases reef aquaria, at somewhat lower than natural salinity levels. This practice stems, at least in part, from the belief that fish are less stressed at reduced salinity. Substantial misunderstandings also arise among aquarists as to how specific gravity really relates to salinity, especially considering temperature effects.

Ron Shimek has discussed salinity on natural reefs in a previous article. His recommendation, and mine as well, is to maintain salinity at a natural level. If the organisms in the aquarium are from brackish environments with lower salinity, or from the Red Sea with higher salinity, selecting something other than S=35 may make good sense. Otherwise, I suggest targeting a salinity of S=35 (specific gravity = 1.0264; conductivity = 53 mS/cm).

Making Standards with Table Salt

Making salinity standards with ordinary table salt requires the ability to make salt solutions of known concentration. The standards given in this article are all best made using accurate weight measurements, both of the salt and the water. Most aquarists, however, do not have access to high quality balances, so volume-based measurements will be provided. Typically, they will not be as accurate as weight-based measurements, but will be adequate for most aquarium purposes.

A recent article in an online culinary magazine suggested that measuring spoons used by cooks are actually a fairly accurate way to measure volumes. In particular, they showed that of the many teaspoons tested, all were within 1% of the standard volume:

"Measuring spoons don't usually get a lot of consideration: bought once and done. But have you ever wondered if your set of spoons is accurate? Would an expensive set do a better job? To find out, the test kitchen purchased 10 different sets of measuring spoons, made from both plastic and stainless steel and ranging in price from $1.99 to $14.99.

We were prepared for large differences in degree of accuracy but found none. All of the spoons weighed in within a few hundredths of a gram of the official standard-not enough to compromise even the most exacting recipe."

Consequently, aquarists can use measuring spoons for measurement of salt volumes. If they have an accurate balance, all the better. In that case, they should just use the mass specified, rather than the volume.

To measure with a measuring spoon or cup that is measured to the top, first overfill it, then use the back of a knife to carefully level the volume. I did this with a variety of different measuring spoons and cups using Morton's Iodized Salt, and got the following results:

• 5 teaspoons = 31.13 g, or 6.2 grams per teaspoon (equivalent to 1.26 g/dry mL)

• 5 tablespoons = 91.04 g, or 18.2 g/tablespoon (equivalent to 1.23 g/dry mL)

• ¼ cup = 73.07 g (equivalent to 1.24 g/dry mL)

• ½ cup = 156.52 g (equivalent to 1.32 g/dry mL)

• 1 cup = 296.62 g (equivalent to 1.25 g/dry mL)
  

In this case, only the ½ cup deviated from a fairly standard value. Overall, the Salt Institute suggests that 1 teaspoon of a variety of different salt brands weighs about 6 grams (1.22 g/dry mL):

"The density of granulated evaporated salt varies depending on crystal size, structure, gradation, and degree of compaction. The reported range of densities is 1,200-1,300 g/L." We will use 1.217 g/mL, which gives 6 grams per teaspoon."

Measuring the water's volume is best done with an accurate measuring container of appropriate dimensions. In the absence of such a container, however, I have measured a container whose volume may be standardized across the United States, and which may therefore allow reasonably accurate volume measurement. In particular, a plastic 2-L Diet Coke bottle filled to the absolute top contained 2104.4 grams (mL) of water. In a pinch, these containers may serve well as volume standards (at least until the company changes bottle styles).

[[Notice added post-publication: the standards described here that use Coke bottles are subject to variation in the volume of a 2-L Coke bottle. It has recently come to my attention that such 2-L bottles can vary in total volume, and that this can lead to at least a 1 ppt error in the salinity of the standards matched to seawater salinity of 35 ppt. Standards made with accurate measurements of salt and water should still accurately match 35 ppt.]]

Refractometer Standard

It is widely believed that only pure water is required to calibrate refractometers. That fact is true of many refractometers, and is certainly appropriate for routine calibration, but it assumes that they were manufactured correctly and have not been damaged since manufacturing. As refractometers used by aquarists become less and less expensive (with some now selling for less than $30), there is every reason to believe that at some point they will no longer be accurate enough.

The only way to be sure that a given refractometer gives useful information is to check its accuracy in a solution similar to aquarium water. I believe that all refractometers should be checked in this fashion when first purchased, and again any time there is a reason to be concerned. For example, an aquarist might be concerned if an aquarium that had been running for years at a salinity of 35 ppt suddenly reads 39 ppt.

In order to provide a standard for refractometers, a solution whose refractive index is similar to normal seawater is required. Seawater with S= 35 has a refractive index of 1.3394.¹ Likewise, the refractive index of different sodium chloride solutions can be found in the scientific literature. My CRC Handbook of Chemistry and Physics (57^th Edition, Page D-252)⁴ has such a table. That table has entries for 3.6 and 3.7 weight percent solutions of sodium chloride that span the value for normal seawater. Interpolating between these data points suggests that a solution of 3.65 weight percent sodium chloride has the same refractive index as S=35 seawater, and can be used as an appropriate standard (Table 2).

This 3.65 weight percent sodium chloride solution can be made by dissolving 3.65 grams of sodium chloride in 96.35 grams (mL) of purified freshwater.

For a rougher measurement in the absence of an accurate water volume or weight measurement:

1. Measure ¼ cup of Morton's Iodized Salt (about 73.1 g)
2. Add 1 teaspoon of salt (making about 79.3 g total salt)
3. Measure the full volume of a plastic 2-L Coke or Diet Coke bottle filled with purified freshwater (about 2104.4 g)
4. Dissolve the total salt (79.3 g) in the total water volume (2104 g) to make an approximately 3.65 weight percent solution of NaCl. The volume of this solution will be slightly larger than the Coke bottle, so dissolve it in another container.

[[Notice added post-publication: the standards described here that use Coke bottles are subject to variation in the volume of a 2-L Coke bottle. It has recently come to my attention that such 2-L bottles can vary in total volume, and that this can lead to at least a 1 ppt error in the salinity of the standards matched to seawater salinity of 35 ppt. Standards made with accurate measurements of salt and water should still accurately match 35 ppt.]]

How to Use a Refractive Index Standard

One simple way to use this refractive index standard is to measure it with a refractometer, and just remember what setting the standard came to. That setting represents S=35 seawater, with all of the properties shown in Table 1. Hopefully, the reading of the refractometer at that point will be similar to the properties in Table 1 (specific gravity = 1.026 - 1.027, or S=35, depending on the units). Simply using it as the target salinity for the aquarium is a fine way to go.

Alternatively, one can actually calibrate the refractometer using the standard by adjusting it until it reads the appropriate setting indicated in Table 1. Exactly how to adjust it depends on the refractometer, but often it is as simple as turning a screw.

Specific Gravity Standard

Most aquarists recognize that inexpensive hydrometers are often prone to error. In some cases, inaccuracy is due to poor manufacturing, and in other cases it is due to poor usage by aquarists. In a previous article I tested several hydrometers and found variable results, from good to marginal. Beyond the inherent accuracy of the measurement is the confusing problem of how specific gravity relates to the temperature of the measurement, an issue which I detailed in that same article.

The best way to be sure that a given hydrometer is giving accurate information is to check its accuracy in a solution with a density (specific gravity) similar to the aquarium water. In order to provide a standard for hydrometers, a solution of a similar specific gravity to normal seawater is required. Seawater with S= 35 has a specific gravity of about 1.0264 (Tables 1 and 3).

In order to match this specific gravity to a standard solution made from sodium chloride, look up the density of different sodium chloride solutions in the scientific literature. My CRC Handbook of Chemistry and Physics (57^th Edition, Page D-252)⁴ has such a table (partially reproduced in Table 4), but it has data only for 20°C (68°F). Specific gravity at 20°C is then easily calculated by dividing the density of the solutions by the density of water at the same temperature. This table (4) can then be compared to seawater at 20°C (Table 5). The primary purpose of showing specific gravity at 25°C (77°F; Tables 1 and 3) and 20°C (Table 4) is to show that specific gravity does not change much with temperature (1.0264 vs. 1.0266). Nevertheless, it is only the 20°C data that will be used to devise a standard.

The table in the CRC Handbook has entries for 3.7 and 3.8 weight percent solutions of sodium chloride that span the specific gravity value for normal seawater. Interpolating between these data points suggests that a solution of 3.714 weight percent sodium chloride has the same specific gravity (and density) as S=35 seawater, and can be used as an appropriate specific gravity standard (Table 5). For most purposes, 3.7 weight percent is accurate enough.

To produce a 3.714 weight percent sodium chloride solution, dissolve 1 teaspoon (6.20 grams) of Morton's Iodized Salt in 161 mL (161 g) of freshwater (making a total volume of about 163 mL after dissolution of the salt). This solution can be scaled up as desired.

For a rougher measurement in the absence of an accurate water volume measurement:

1. Measure ¼ cup of Morton's Iodized Salt (about 73.1 g)
2. Add 1½ teaspoon of salt (making about 82.4 g total salt)
3. Measure the full volume of a plastic 2-L Coke or Diet Coke bottle filled with purified freshwater (about 2104.4 g)
4. Add an additional 2 tablespoons of purified freshwater (about 30 g)
5. Dissolve the total salt (82.4 g) in the total water volume (2134.4 g) to make an approximately 3.7 weight percent solution of NaCl. The volume of this solution is larger than the Coke bottle, so dissolve it in another container.

[[Notice added post-publication: the standards described here that use Coke bottles are subject to variation in the volume of a 2-L Coke bottle. It has recently come to my attention that such 2-L bottles can vary in total volume, and that this can lead to at least a 1 ppt error in the salinity of the standards matched to seawater salinity of 35 ppt. Standards made with accurate measurements of salt and water should still accurately match 35 ppt.]]

How to Use a Specific Gravity Standard

Depending on the type of hydrometer, one would use this solution differently.

For standard floating hydrometers (Figure 2), which are not self-correcting for temperature variations, it is important to use the standard at the same temperature at which the aquarium water will be tested (within say, ± 0.5 °C or ± 1 °F). Preferably, that will also be the temperature at which the hydrometer is intended to be used (often marked on it), but that is not an absolute requirement. The aquarist can then mark on the hydrometer the level to which it rises (that is, the water line), and use that as an indication of the specific gravity of S=35 seawater, which has all of the properties listed in Table 1(specific gravity = 1.0264, etc). If the hydrometer reads higher or lower than 1.0264, then the aquarist can just imagine the scale on the hydrometer to be shifted up or down, and shift all other readings taken with it (at the same temperature) by the same amount.

For example, if the standard comes out at 1.0230 (and it is really 1.0264), then just add 1.0264 - 1.0230 = 0.0034 to each measured value).

For swing arm hydrometers (Figure 3), which are largely self-correcting for temperature variations, add the standard to the swing arm hydrometer at roughly the same temperature at which the aquarium water will be tested (say, ± 5°C or ± 10°F). Once the reading stabilizes, the aquarist can mark the reading (or just remember it) and use that as an indication of the specific gravity of S=35 seawater, which has all of the properties listed in Table 1 (specific gravity = 1.0264, etc). If the hydrometer reads higher or lower than 1.0264, then the aquarist can just imagine the scale on the hydrometer to be shifted up or down, and shift all other readings taken with it by the same amount, just as for a standard floating hydrometer.

Just to be especially clear: this solution need not be used at exactly 20°C (68°F). It will be just about as accurate at 25°C (77°F) since specific gravity does not change much with temperature, and these salt solutions would be expected to change density with temperature in about the same fashion as seawater. The most important factor is that the temperature of the standard, when measured, be the same as the aquarium water when it is measured..

How to Use a Standard Hydrometer

Here are a few additional tips for using a hydrometer:

1. Make sure that the hydrometer is completely clean (no salt deposits) and that the part of the hydrometer above the water line is dry. Tossing it in so it sinks deeply and then bobs to the surface will leave water on the exposed part that will weigh down the hydrometer and give a falsely low specific gravity reading. Salt deposits above the water line will have the same effect. If any deposits won't easily dissolve, try washing it in dilute acid (such as vinegar or diluted muriatic acid).
2. Make sure that there are no air bubbles attached to the hydrometer. These will help buoy the hydrometer and yield a falsely high specific gravity reading.
3. Make sure that the hydrometer is the same temperature as the water (and preferably the air).
4. Read the hydrometer at the plane of the water's surface, not along the meniscus (Figure 2; the meniscus is the lip of water that either rises up along the shaft of the hydrometer, or curves down into the water, depending on the hydrophobicity of the hydrometer).
5. Rinse with purified freshwater after use to reduce deposits.
6. Do not leave the hydrometer floating around in the tank between uses. If left in the aquarium, deposits may form that will be difficult to remove.

How to Use a Swing Arm Hydrometer

In addition to those described above, here are some special tips for swing arm hydrometers:

7. Make sure that the hydrometer is completely level. A slight tilt to either side will change the reading.
8. Some swing arm hydrometers recommend "seasoning" the needle by filling it with water for 24 hours prior to use. This presumably permits the water absorbed into the plastic to reach equilibrium. In the case of the hydrometer that I tested in a previous article, the hydrometer became slightly less accurate after "seasoning."

Conductivity Standard

Conductivity can readily be used to measure the salinity of seawater. In a previous article I detailed how this measurement works and why it is suitable for reef aquaria. In short, the more ions there are in solution, the more easily the solution will conduct electricity. In fact, conductivity is so easily measured and standardized that it forms the basis of the modern definition of salinity, PSU (Practical Salinity Units). S=35 seawater is defined as seawater with the same conductivity as a solution made from 3.24356 weight percent potassium chloride (KCl), and that conductivity is exactly 53 mS/cm (mS/cm is one of the units used for conductivity, it is milliSiemens per centimeter). Higher and lower conductivities give higher and lower salinities, respectively, using a complicated equation that will not be discussed here.

There are two ways to formulate a conductivity standard that matches S=35 seawater. The first is looking to the scientific literature to see what sodium chloride solutions provide a conductivity of 53 mS/cm. A second is to match a sodium chloride solution to the conductivity of 3.24 weight percent potassium chloride in water. This article does both.

First, the scientific literature. Fortunately, many measurements of conductivity of such solutions have been made over the years. Without going into detail about how they were measured, the data from these papers indicate that a 53 mS/cm conductivity solution is provided by a 33.64 g/L (0.576 M) sodium chloride solution. That solution corresponds to 3.29 weight percent sodium chloride.²

Alternatively, one can measure conductivity of salt solutions. I made a solution of 3.24 weight percent KCl in deionized water and measured its conductivity. The reading on the uncalibrated meter was 52.5 mS/cm (it would have been 53 mS/cm with a perfectly calibrated meter).

I then made a solution of deionized water and Morton's Iodized Salt, adding salt until I matched the conductivity of the prior solution. It required EXACTLY 3.29 weight percent sodium chloride to match this conductivity. Believe it or not, I didn't even recognize the close agreement between these two methods during the test, as I hadn't worked through the math until long after taking the original measurements.

So not only is there good evidence that a 3.29 weight percent sodium chloride solution is appropriate, but additional evidence demonstrates that Morton's Iodized Salt from a grocery store is a suitable material for this purpose.

To make a 3.29 weight percent sodium chloride solution, dissolve 1 teaspoon (6.20 grams) of Morton's Iodized Salt in 182 mL (182 g) of freshwater (making a total volume of about 184 mL after dissolution of the salt). This solution can be scaled up as desired.

For a rougher measurement in the absence of an accurate water volume measurement:

1. Measure ¼ cup of Morton's Iodized Salt (about 73.1 g)
2. Measure the full volume of a plastic 2-L Coke or Diet Coke bottle filled with purified freshwater (about 2104.4 g)
4. Add 3 tablespoons of purified freshwater (about 45 g)
5. Dissolve the total salt (73.1 g) in the total water volume (2149.4 g) to make an approximately 3.29 weight percent solution of NaCl. The volume of this solution is larger than the Coke bottle, so dissolve it in another container.

[[Notice added post-publication: the standards described here that use Coke bottles are subject to variation in the volume of a 2-L Coke bottle. It has recently come to my attention that such 2-L bottles can vary in total volume, and that this can lead to at least a 1 ppt error in the salinity of the standards matched to seawater salinity of 35 ppt. Standards made with accurate measurements of salt and water should still accurately match 35 ppt.]]

How to Use a Conductivity Standard

How to best use a conductivity standard depends a bit on the meter involved. If the meter can be calibrated, then my suggestion is to get the solution to about 25°C (exactly that temperature if the meter doesn't automatically compensate for temperature, but that would be unusual) and then adjust the meter until it reads 53 mS/cm or S=35 (depending on the output).

Many meters, however, do not allow such calibration. In that case, measure the conductivity or salinity of the standard, and then set up a correction ratio that is applied manually. For example, if the standard reads 56 mS/cm, then multiply all readings on that meter by 53/56 (0.946) to get a corrected reading. The same correction could apply to salinity. For example, if it reads S=38 (or 38 ppt), then multiply every reading by 35/38 = 0.921.

Alternatively, the simplest way is to use the value that is found from the standard as the target for the aquarium, and not worry about calibrations or corrections.

Summary

This article provides a way for reef aquarists to make and use salinity standards for the most common ways of measuring salinity: refractometers, hydrometers, and conductivity meters. Hopefully, these will help aquarists avoid problems that might arise from poorly calibrated devices, or at least ease their concerns about whether or not their devices are working properly.

Happy Reefing!

References:

1. The data on the refractive index of seawater as a function of salinity were obtained from:

Practical Handbook of Marine Science, Third Edition Michael J. Kennish; Editor (2000) 896 pp, Publisher: Lewis Publishers, Inc.

2. The data on the conductivity of sodium chloride solutions were obtained from:

Conductances of concentrated aqueous sodium and potassium chloride solutions at 25°C Chambers, J. F.; Stokes, Jean M.; Stokes, R. H. Univ. W. Australia, Nedlands, J. Phys. Chem. (1956), 60 985-6.

3. Data for the physical properties of seawater (other than refractive index) came from a calculator on the Ocean Teacher web site of the Intergovernmental Oceanographic Commission (IOC):

See the new Rite Aid ad for pharmacy coupons and deals.

http://ioc.unesco.org/oceanteacher/~Calculator.htm

4. CRC Handbook of Chemistry and Physics. 57^th ed. Weast, Robert C.; Editor. (1976), 2400 pp. Publisher: (Chem. Rubber Co., Cleveland, Ohio).