Sandfish Hatchery Water Quality, TAN and Oxygen

Sandfish hatchery water-quality logs do not become useful simply because their columns are full. A reading of total ammonia nitrogen, pH, salinity, temperature or dissolved oxygen is a measurement made in a particular tank at a particular time, not an automatic verdict about the health of Holothuria scabra. The interpretation begins with what was sampled, which chemical quantity a method measures, and what decision the operator actually needs to make.

This guide translates the technical language of sandfish hatchery water quality into a reproducible method for reading logbooks and laboratory reports. Its central distinction is between total ammonia nitrogen, often abbreviated TAN, and the fraction present as un-ionised ammonia, NH3. A second distinction separates dissolved oxygen concentration in milligrams per litre from percent saturation, which depends on temperature, salinity and pressure.

Worked figures in this article are explicitly hypothetical teaching examples, not measurements from a real hatchery and not recommended species-specific tolerance limits. They show how to preserve units, recognise missing evidence, check a calculation and pass an intelligible result to the next technician without inventing certainty.

Start with a tank, a sample and a decision

A good log begins with the identity of the culture unit, biological stage, collection depth, time and instrument. The same number can describe a broodstock tank, a larval rearing vessel or a juvenile nursery, but each has different loading, feeding, microbial and water-exchange patterns. If a result arrives without the vessel and stage, a reader cannot know whether it belongs to today’s feeding event, yesterday’s partial exchange or another system altogether. A useful interpretation first constructs a timeline of observations rather than treating a spreadsheet column as an independent story.

Ask a narrow decision question. Is the purpose to verify an instrument alarm, investigate mortality, decide when to resample, understand a post-feeding shift, or compare two water sources? A result can be technically valid while inadequate for a particular decision. Recording the question makes the requested evidence visible: a one-off midday pH reading will not establish an overnight oxygen minimum, and a laboratory TAN result without a matched pH and temperature will not establish the NH3 fraction at the sampling moment.

Know what the species-specific literature can and cannot establish

Holothuria scabra hatchery practice is documented in the SEAFDEC Aquaculture Department’s sandfish production manual, which follows broodstock, algal cultivation, larvae, settlement and juvenile handling. Its practical sequence is useful for identifying where water-quality measurements belong in the workflow. It should not be turned into a universal statement that a single ammonia, oxygen or salinity value is safe in every facility. Culture stage, local water, acclimation, measurement quality and the duration of exposure can alter the implications.

Other aquaculture water-chemistry documents explain the general NH3/NH4+ equilibrium and oxygen measurement, but their toxicity endpoints may concern fish or different invertebrates. Transfer the chemistry, not an unsupported species threshold. Whenever a report uses a generic aquaculture chart, describe exactly what is transferable and what remains unverified for sandfish. This distinction is a core translation skill: scientific agreement about chemical form does not imply agreement about an organism’s response under every experimental condition.

Build a defensible sampling record

At minimum, give the tank identifier, clock time and timezone, date, sampled position and depth, sample type, temperature, salinity, pH, dissolved oxygen, ammonia method and units, and the operator or laboratory. Note whether water was aerated during sampling, whether a net or animal was handled immediately before collection, and whether a water exchange had just occurred. A log with an exact three-decimal result but no sampling context suggests precision it cannot defend. The missing context can be more consequential than the number of decimal places.

Pair chemistry results from the same water and time window whenever possible. An ammonia sample couriered to a laboratory and a pH reading recorded hours later can describe two different chemical situations, even if both receive the same date in a spreadsheet. If only unmatched measurements exist, mark that limitation rather than silently combining them. Document sample preservation, holding time and instrument calibration because numbers change meaning when collection and analytical procedures differ.

Read TAN as a measured nitrogen pool, not as pure NH3

Total ammonia nitrogen includes the nitrogen present in both dissolved ammonium ion, NH4+, and un-ionised ammonia, NH3. A common reporting form is milligrams of nitrogen per litre, written mg N/L or mg/L as N. That nitrogen basis matters: an ammonia molecule weighs more than the nitrogen atom it contains. A laboratory may also report mg/L NH3, mg/L NH4+ or a kit-dependent equivalent. Do not subtract or compare numbers unless the reporting species and basis are explicitly aligned.

A sample labelled TAN = 0.20 mg N/L does not state that 0.20 mg of free NH3 exists in every litre. It describes combined ammonia nitrogen. To translate it into the un-ionised fraction, the reader needs the applicable equilibrium calculation or a validated table that accounts for the relevant water composition, temperature and pH. Calling all TAN ‘toxic ammonia’ overstates one chemical form; calling ammonium harmless under all circumstances is also unjustified.

Why pH changes the chemical question

Ammonium and un-ionised ammonia are connected by an acid-base equilibrium. As pH rises under otherwise comparable conditions, a larger proportion of the TAN pool is present as NH3. The practical consequence is that two samples can show the same TAN but a different NH3 fraction. An operator who watches TAN alone can miss an important change when daytime photosynthesis, carbon dioxide exchange or another process raises pH. The point is not to assume a particular risk threshold, but to know which quantity is actually changing.

The pH scale is logarithmic; a one-unit rise is a tenfold reduction in hydrogen-ion activity under the formal definition, not a simple one-percent movement. A pH reading should therefore be interpreted alongside instrument uncertainty, temperature compensation and calibration history. If one log says pH 8.0 and another says 8.6, the correct reaction is to investigate the timing and measurement conditions before calculating a shared NH3 value from either.

Temperature and salinity are inputs, not decoration

Temperature alters the ammonia equilibrium and the biological context of exposure. Salinity and ionic strength matter too, particularly in marine and brackish water. A freshwater lookup table may not reproduce the correct NH3 fraction in seawater, even if the pH and temperature columns appear conveniently matched. The US Geological Survey’s discussion of ionic-strength corrections explains why transferring freshwater ammonia calculations to saline water can introduce meaningful error. The governing table or model should identify its water chemistry assumptions.

A report with TAN, pH and temperature but no salinity can still describe its measured quantities; it may not support a precise marine NH3 fraction. In that situation, state a conditional estimate only if the assumptions are openly specified and suitable. If a decision depends on a small difference, obtain the missing measurement and use a marine-appropriate calculation rather than padding the report with digits. Good translation preserves the boundary between an observation and a modelled result.

Keep NH3-N separate from molecular NH3 mass

When an un-ionised fraction f is applied to TAN reported as mg N/L, the product is NH3-N in mg N/L. The numerator remains nitrogen mass even though that nitrogen is assigned to the NH3 chemical form. If a stakeholder instead needs mg NH3/L, multiply NH3-N by the molecular-mass ratio 17/14, approximately 1.214. That conversion does not change the amount of nitrogen; it changes which mass is being named. Confusing the two conventions creates a predictable twenty-percent-scale reporting error.

For a deliberately hypothetical calculation, let TAN equal 0.20 mg N/L and suppose an appropriate speciation calculation gives f = 0.06 at the measured conditions. The corresponding NH3-N is 0.012 mg N/L. Molecular NH3 concentration is about 0.0146 mg NH3/L. The fraction 0.06 here is a supplied example input, not an experimentally verified fraction for a particular sandfish tank. A report must state where its actual fraction came from.

A fraction must be a fraction, not a percentage number

A model may return an un-ionised fraction of 0.04, while a reference table may display 4%. These are the same share, but multiplying TAN by 4 instead of 0.04 creates a hundredfold error. The safe calculation writes the unitless factor before arithmetic: 4% = 4/100 = 0.04. This matters when a spreadsheet cell is formatted as a percentage but a pasted value is interpreted as a plain number. Always inspect underlying cell values as well as their visible labels.

In a checkable example, TAN is 0.35 mg N/L and a validated method produces 4%. Multiplying 0.35 by 0.04 gives 0.014 mg NH3-N/L. Multiplying by 4 would give 1.4 mg N/L, greater than the entire TAN pool; that violates the basic relation that a fraction of a pool cannot exceed the pool when f lies between zero and one. Dimensional and magnitude checks can catch this mistake before a flawed chart reaches an operator.

Do not import a fish toxicity threshold into a sandfish chart

Fish-related ammonia tables may have a legitimate purpose in fish farming or water-quality regulation, but they are not automatically biological action limits for sea cucumber larvae or juveniles. A threshold depends on species, stage, pH, temperature, exposure duration, endpoint and study quality. Even within one species, survival after a short exposure is not synonymous with no sublethal effect on feeding, settlement or later growth. The right interpretation states what the source studied and declines to invent a missing sandfish-specific limit.

If a hatchery has a locally validated action plan, quote its current documented limit, the measured chemical basis and the response procedure together. Otherwise describe whether an observation is above the facility’s ordinary baseline, whether it is changing unusually fast, and what additional expert assessment is needed. A trustworthy article can teach chemistry without pretending to prescribe animal-health or water-treatment decisions from one synthetic example.

Dissolved oxygen concentration is a different unit family

Dissolved oxygen, abbreviated DO, is commonly measured in mg O2/L. Unlike TAN reported as nitrogen mass, DO expresses molecular oxygen mass in water volume. A recorded 5.1 mg/L is a concentration, not a percentage of air and not a measure of the amount of air pumped by an aerator. Operators also use percentage saturation, which compares measured DO with a reference solubility at the relevant temperature, salinity and barometric conditions. The two forms answer related but non-identical questions.

At a fixed measured DO, percent saturation can change if the expected solubility changes. Warm water generally holds less equilibrium oxygen than cooler water, and salinity also changes solubility. USGS DOTABLES explicitly takes temperature, pressure and salinity as inputs when determining saturation or solubility values. Do not copy a freshwater saturation percentage into saline hatchery water. State which reference was used and whether the instrument’s compensation settings were correctly configured.

Check oxygen saturation with a transparent denominator

Percent saturation is 100 times measured DO divided by the applicable equilibrium DO concentration. Suppose an illustrative instrument observes 4.55 mg O2/L and a separately calculated reference for those specific conditions is 6.50 mg O2/L. The percentage is 100 × 4.55/6.50 = 70%. This is a mathematical example, not a claim that 6.50 mg/L is the equilibrium value for any named site or that 70% is an acceptable sandfish threshold. The required denominator changes with environmental conditions.

A figure above 100% can be physically meaningful in supersaturated water, for example during strong photosynthetic oxygen production; it is not automatically a broken sensor. Equally, a surprising value merits checks for calibration, bubbles on the probe, sensor membrane condition or compensation configuration. A translator should not erase a seemingly inconvenient result simply because it falls outside the intuitive 0–100 range. The number needs its physical context before it earns an interpretation.

The time of day can matter more than a daily average

Oxygen and pH are dynamic, particularly where photosynthesis and respiration vary between daylight and darkness. A single noon sample can miss a pre-dawn oxygen minimum and can represent a different pH context from a nighttime TAN sample. To judge exposure duration, a continuous or appropriately repeated log is more informative than an average computed from a few convenient readings. The schedule should reflect the biological question and the facility’s validated monitoring plan rather than an arbitrary clock.

Suppose three illustrative DO values are 6.2, 5.3 and 3.1 mg/L at different times. Their arithmetic mean is about 4.87 mg/L, but that average does not disclose how long water remained near the lowest observation. Neither does the minimum establish duration without a sampling interval. Translating these figures responsibly requires a chart of timestamps, possible sensor gaps and the circumstances surrounding the decline. An average is a summary, not a substitute for exposure history.

Tank inventory translates concentration into mass

A concentration is a ratio. To estimate the total amount of TAN-N dissolved in a well-mixed tank, multiply concentration in mg N/L by volume in litres. A 4,000-L hypothetical tank with TAN 0.20 mg N/L contains 800 mg, or 0.80 g, of dissolved TAN-N at the sampling moment. This is an inventory estimate, not a statement of daily ammonia production or of the amount absorbed by animals. It assumes the stated volume and sample represent the relevant body of water.

If the tank actually holds 3,600 L because water was drained, using its 4,000-L nominal capacity overstates inventory by about eleven percent relative to the true 720-mg amount at 0.20 mg N/L. Note displacements, working water level, dead zones and mixing where they materially affect the result. A precise TAN test does not rescue an inaccurate volume or an unrepresentative sample. Distinguish tank inventory from ammonia loading rate in every subsequent calculation.

Why water exchange is not a magic dilution button

For ideal instantaneous mixing and no production or removal during the event, a partial replacement can be modelled as Cafter = (1-r)Cbefore + r Csource, where r is the volume fraction exchanged. For illustrative numbers, take 20% of a tank at TAN 0.30 mg N/L and replace it with water at 0.05 mg N/L. The model predicts 0.8 × 0.30 + 0.2 × 0.05 = 0.25 mg N/L. This is the expected mixed concentration under the stated assumptions, not a guaranteed post-exchange observation.

Real exchanges may be continuous, stratified or accompanied by altered feeding, microbial activity, sediment disturbance or source-water variability. The exchange volume may be an estimate and the source may contain ammonia. If the measured post-exchange TAN is 0.29 mg N/L, first verify actual water volume, complete mixing, sampling time and source concentrations. Do not decide that the laboratory was wrong simply because a simplified model predicted a different outcome. The model explains a baseline expectation and helps identify what to investigate.

How to distinguish production rate from water inventory

Ammonia nitrogen may enter via source water, excretion or decomposition and leave through exchange, biological transformations or other sinks. The difference between two tank inventories therefore cannot simply be labelled excretion. If inventories rise by 400 mg N between samples, that is a net change across the measured intervals under the assumed volumes. The gross production might have been larger if nitrogen was simultaneously removed; it might also reflect unmeasured inflow. A mechanistic nitrogen budget needs defined boundaries and measured fluxes.

For example, if inventory rises from 0.8 g to 1.1 g while 0.2 g N enters from source water and 0.1 g N is removed with discharged water, the other net processes together account for an increase of 0.2 g N. That residual does not individually identify excretion, decomposition or nitrification. Record uncertainty in the inflows and outflows rather than forcing a residual into a biological label. Translation should preserve what a balance can infer and what requires separate evidence.

Nitrification changes species, not simply the word ‘ammonia’

Microbial processes can convert ammonia nitrogen into nitrite and then nitrate under appropriate conditions. Thus decreasing TAN does not automatically mean total dissolved inorganic nitrogen has disappeared from the system. It may have changed chemical form. Nitrite and nitrate reports use their own nitrogen-basis or molecular-ion conventions. If a log says nitrite = 0.10 mg/L, it is essential to know whether the laboratory means NO2-N or NO2- ion before adding concentrations into any nitrogen budget.

The nitrogen mass fraction in nitrite is approximately 14/46; therefore 0.10 mg NO2-/L corresponds to about 0.0304 mg NO2-N/L. For nitrate, 14/62 gives about 0.0226 mg NO3-N/L from 0.10 mg NO3-/L. These are arithmetic conversions, not evidence that a sandfish tank has reached an acceptable condition. A combined nitrogen figure should use consistent nitrogen units and identify which measured species were included or omitted.

Read salinity units and instrument conditions carefully

Salinity can appear in practical salinity units, a practical salinity value, parts per thousand or grams per kilogram, depending on the instrument and convention. These are not always interchangeable without qualification. A conductivity probe may estimate practical salinity based on calibration and assumptions about seawater composition. Water warmed during collection can produce a different raw conductivity reading even when a temperature-compensated salinity output is relatively stable. Record the actual displayed parameter and device settings.

Some reports simply place ’30 ppt’ beside a sample. Ask whether this is a refractometer value, a gravimetric mass fraction or a field display convention. Two technicians using different instruments may obtain superficially matching numbers that carry different uncertainties. Salinity is also a variable in marine NH3 calculations and oxygen solubility. If it is absent or ambiguously defined, annotate downstream estimates accordingly instead of filling in a locally typical value and treating that assumption as observed fact.

Carbon dioxide and pH should be interpreted together

Respiration and gas exchange influence dissolved carbon dioxide and thus the carbonate system in seawater. pH alone does not uniquely determine alkalinity, dissolved inorganic carbon or carbon dioxide partial pressure. A report that uses only a pH column to declare an oxygen or carbon dioxide problem is overinterpreting the evidence. In hatchery tanks, gas exchange, biological activity and water replacement can shift different aspects of the system at different speeds.

For a useful investigation, match pH with temperature, salinity, time, any alkalinity or carbon-system measurements and a description of aeration and water exchange. An increase in pH could coincide with different changes in DO depending on the process. Translating the log means refusing a simplistic rule such as ‘higher pH means better water’. The right question is which process could plausibly explain the joint observations and what independently measurable prediction would test it.

Measurement uncertainty has several sources

Instrument resolution is not the same as measurement accuracy. A DO sensor displaying 5.23 mg/L may have calibration uncertainty, response lag, fouling or compensation errors larger than its final decimal digit. Laboratory TAN results have their own precision, detection limits, recovery checks and blank corrections. Sampling introduces further variation across tanks, depths or times. A responsible report identifies major uncertainty contributors rather than turning a string of decimal places into a claim of certainty.

Where replicate measurements exist, report their scatter and how replicates were collected. Repeated readings from the same stationary probe estimate short-term instrumental variability, not spatial variation throughout a heterogeneous vessel. Replicate bottles taken at multiple locations may better address distribution but add sampling error. If uncertainty is unknown, do not fabricate a confidence interval. State ‘uncertainty not supplied’ and name the missing calibration, replicate or laboratory information necessary to estimate it.

Detection limits and censored values require special care

A laboratory entry such as TAN less than 0.02 mg N/L means the concentration was not reported as a quantified value above the stated threshold. It does not equal zero, and it should not be replaced automatically with 0.01 mg N/L without an explicit statistical or analytic convention. Likewise, ‘not detected’ does not prove that ammonia nitrogen or any other analyte is absent. The detection or reporting limit, method and sample conditions govern what that phrase means.

If a sequence contains 0.03, less than 0.02 and 0.04 mg N/L, computing an ordinary mean by entering the middle result as zero may bias the trend. In a quick operational table, preserve the censored notation and describe that a single numerical average cannot be obtained without an assumption. For formal analysis, choose a method suitable for censored data and the purpose of the inference. The purpose of translation is to keep the laboratory’s limit visible, not to hide it.

Temperature compensation is part of the measurement

A pH electrode reports an electrical potential affected by temperature. Its automatic temperature compensation commonly addresses electrode response; it does not guarantee that the chemical solution’s pH is unchanged by temperature. Dissolved oxygen meters may use membrane, optical or other measurement principles with compensation and calibration steps. Calibration under one condition followed by measurement under another can create avoidable differences if configuration is wrong.

Inspect the instrument’s calibration date, standards, salinity setting, barometric setting where relevant and maintenance notes before arguing that a small numerical change represents biology. A probe held near bubbles may behave differently from one placed in a well-mixed region, and a sensor allowed insufficient time to stabilise can report a transient. A defensible translation says ‘observed after compensation as configured’ rather than claiming the device has measured an absolute biological truth.

Maintain a change log for the water system

A chemical trend becomes much easier to explain when the operational record includes feeding, added microalgae, cleaning, tank transfers, biofilter service, aeration changes, temperature-control failures and source-water replacements. Each event has a start time, end time and magnitude. Without that record, a reader may mistake a deliberate intervention for spontaneous improvement. For example, lower TAN after a large exchange cannot be assigned confidently to improved biological assimilation.

Create separate columns for event timing, amount and location rather than writing ‘tank fixed’ in a free-text cell. If source-water chemistry changes, record a paired source sample. If a pump briefly stops, note the affected circulation path and exact duration. These facts permit later reconciliation with oxygen and nitrogen measurements. A good technical narrative reconstructs which processes were possible at the moment of each sample, rather than offering a plausible but untestable anecdote.

Do not confuse stock density with water quality causation

High animal numbers can influence loading, but density by itself does not quantify ammonia production or oxygen consumption. It must be connected to organism size, stage, feeding, microbial activity, flow, water volume and time. If a 1,000-L tank carries 1,000 juveniles, the count density is one per litre; it is not a mass loading rate. Another tank carrying the same count of much larger animals may have very different metabolic and particulate inputs.

Compare an animal inventory based on a documented count and size distribution with a contemporaneous water chemistry and operating record. A change in stocking density may coincide with a cleaning event, new feed, different substrate or temperature shift. A correlation among such records should not be presented as a demonstrated causal relationship unless alternative explanations have been examined. This preserves the practical value of density data without turning it into a false universal predictor.

Read alarms as decisions with defined triggers

An alarm rule may use absolute concentration, rate of change, an instrument status flag, a time-window average or a validated local biological action limit. Those are different logical conditions. ‘Alert if DO falls below X’ is not equivalent to ‘alert if DO decreases by X in an hour’. A notification can also be triggered by missing data rather than a dangerous measured condition. The report must identify the rule version and whether the instrument was functioning when the trigger occurred.

Where no locally approved biological limit exists, a method can still flag an unexpected deviation for verification and expert review. But it should avoid presenting a generic fish standard as a sandfish safety certificate. A practical handover distinguishes ‘measurement confirmed’, ‘change under investigation’, ‘sample pending’ and ‘action chosen by responsible operator’. That vocabulary prevents the dangerous compression of observation, interpretation and intervention into one red or green symbol.

Worked case: a rise in pH with constant TAN

Consider hypothetical samples at 09:00 and 15:00. TAN remains 0.25 mg N/L in both; pH rises from 7.9 to 8.4 while temperature and salinity are measured and documented. A suitable marine speciation calculation might return different fractions; for a purely illustrative arithmetic demonstration, suppose f changes from 0.03 to 0.08. Then NH3-N moves from 0.0075 to 0.020 mg N/L despite unchanged TAN. The factor increase is about 2.67. The figures for f are scenario inputs, not universal values for those pH readings.

The correct conclusion is that chemical speciation can change without total pool concentration changing. The example does not establish toxicity or predict animal survival. To act on a real log, the operator must calculate fractions appropriate to the actual chemistry, verify analytical timing and compare the findings with facility-specific evidence. If pH was taken after a partial exchange while TAN was from before it, the neat two-time calculation would not be justified.

Worked case: apparent oxygen improvement after an instrument change

Suppose one logger averages 4.8 mg O2/L before replacement, and another reports 5.5 mg O2/L afterward. It is tempting to describe a 0.7-mg/L improvement. But the meters may use different salinity compensation, calibration or placement. A comparison cannot attribute the difference to biological or operational improvement unless the instruments are demonstrably comparable under a bridging test and the tank conditions are controlled or documented.

An informative audit would measure a common sample with both instruments, ideally across relevant oxygen levels, and compare their readings with an independently maintained reference or accepted calibration procedure. If the old meter reads 4.8 and the new 5.3 on the same water, a substantial component of the historical jump could be instrument-related. Report the overlap results and limitations rather than silently splicing two series into one continuous trend.

Worked case: oxygen mass and water volume

A hypothetical 2,500-L tank reads 5.0 mg O2/L. If well mixed, the dissolved oxygen inventory is 2,500 times 5.0 = 12,500 mg O2, or 12.5 g. That number is a momentary inventory; it is not the tank’s daily oxygen supply. Aeration, photosynthesis, respiration, exchange and surface transfer can replenish or remove oxygen continuously. A large inventory alone does not establish a reliable oxygen delivery rate.

If water volume falls to 2,000 L while concentration stays at 5.0 mg/L, the inventory becomes 10 g, even though the instrument concentration is unchanged. Conversely, 12 g in 2,000 L means a 6.0-mg/L concentration under uniform mixing. This example illustrates why tank-level quantity and concentration must not be merged. It does not imply an operational recommendation about how much oxygen a sandfish hatchery should contain.

Worked case: separate a chart error from a true anomaly

Imagine a graph in which TAN suddenly climbs from 0.15 to 15 mg N/L. Before concluding a hundredfold biological event, check whether one result was entered as micrograms per litre and another as milligrams per litre, whether the decimal separator changed, or whether the second entry is reported as ammonia molecule mass instead of nitrogen mass. Unit conversions can create large artificial discontinuities. This does not mean such a large rise is impossible; it means the evidence must be reconciled.

Keep the original laboratory report, instrument export and transformed spreadsheet value as distinct fields. A corrected chart should retain an audit trail showing the original entry, identified problem, conversion rule, person checking it and date. If no explanation is found, preserve the high observation as unverified and investigate with another sample as appropriate. Quietly deleting uncomfortable outliers makes later interpretation less trustworthy.

Common language mistakes in translations and handovers

Translating TAN as ‘ammonia gas’ changes the measured chemical quantity. Translating NH3-N as ‘NH3’ without the nitrogen basis changes its mass interpretation. Calling percent oxygen saturation a ‘percent oxygen concentration’ blurs a ratio and an amount. Referring to salinity in practical salinity units as simply a mass percentage can also cause misleading arithmetic. A technical glossary should map each original term to its meaning, unit and measurement basis, not merely a familiar-looking word.

For multilingual teams, preserve standard symbols TAN, NH3, NH4+, DO, mg N/L, mg O2/L, pH and PSU-like salinity notation where appropriate, and explain each in the reader’s language. Do not translate a laboratory’s method name into a different method. A clear handover can say that TAN was measured, the NH3 fraction was calculated using documented conditions, and oxygen saturation was obtained with a specified correction. That sentence protects the distinction between measurements and derived estimates.

How to write an honest summary paragraph

An effective technical summary begins with observations: which tank was monitored, over which interval, with what methods and units. It then explains derived values and uncertainty: for example, a marine-appropriate ammonia fraction calculation used paired TAN, pH, temperature and salinity. Only after that does it discuss possible causes, alternatives and the decision made by the responsible team. This order prevents a plausible hypothesis from appearing to be a laboratory fact.

One useful reporting structure is ‘What was observed; what was calculated; what the result does not establish; what would test the explanation.’ A long report is not automatically more reliable than a short one, but explicit boundaries matter. If a sensor failed for part of the night, include the gap prominently. If no threshold specific to the developmental stage was available, do not imply compliance. A reader should be able to reproduce the math and see the limits of interpretation.

Practice problem one: reconcile two ammonia conventions

A laboratory gives TAN = 0.28 mg N/L; a valid calculation for the sampled saline water returns an un-ionised fraction of 5%. Find NH3-N and molecular NH3 concentrations, then explain whether the result establishes safety. Convert 5% to 0.05 and multiply: 0.28 × 0.05 = 0.014 mg NH3-N/L. Apply 17/14 to obtain about 0.017 mg NH3/L. Neither result, standing alone, establishes species- or stage-specific safety because exposure duration and a validated biological reference are not supplied.

Challenge the result with a magnitude test. NH3-N must not exceed the total nitrogen pool represented by TAN if the factor is a fraction between zero and one. If an answer returns 1.4 mg NH3-N/L, the factor was almost certainly entered as 5 instead of 0.05. The correction is mathematical, not a new laboratory observation. Preserve the original TAN value, identified speciation method and calculated mass convention in the report.

Practice problem two: interrogate a dissolved oxygen percentage

A report states DO = 4.2 mg/L and saturation = 75%. What equilibrium concentration does that imply? Rearranging the ratio gives 4.2/0.75 = 5.6 mg/L. The reader should verify whether 5.6 mg/L is a defensible reference for the logged temperature, salinity and pressure using the declared calculation or sensor configuration. It is not legitimate to assume that a textbook freshwater reference automatically applies to this saline sample.

Suppose the original report instead lists 4.2 mg/L and saturation = 750%. That mismatch prompts a check for a decimal-place error, percentage-format error, or an unusual physical situation before correction. It does not permit automatic replacement with 75% without inspecting raw instrument output. A translator’s role is to flag and test the ambiguity, not rewrite the record to create a more ordinary-looking graph.

Practice problem three: interpret a mixed-tank dilution

A hypothetical tank has TAN 0.40 mg N/L. Thirty percent of its working water volume is replaced by source water at TAN 0.10 mg N/L. Assuming ideal mixing and no other transformations, the prediction is 0.70 × 0.40 + 0.30 × 0.10 = 0.31 mg N/L. The result is not 0.28 mg N/L because the replacement water is not ammonia-free. Show both contributions in the written calculation so that a reviewer can identify what was assumed.

If a real post-exchange test reports 0.35 mg N/L, compare the sampling interval, mixing, exact exchanged fraction and other nitrogen sources or sinks. Do not force a single mechanistic explanation. An uncertainty in the true exchange fraction or differences in source chemistry could affect the prediction. The correct answer includes the computed ideal result and the conditions under which that result might fail.

Practice problem four: convert a nitrogen inventory

A 3,200-L tank contains TAN 0.18 mg N/L. The estimated pool is 576 mg N, or 0.576 g N. If a 400-L volume change happens before the next sample, the calculation must use the new actual volume; a new concentration alone will not tell how much TAN mass left in the discharged water without knowing the discharge concentration and mixing conditions. The result cannot be described as a rate because the observation has no time denominator.

Now suppose the concentration after an intervention is 0.15 mg N/L in 2,800 L. The new inventory is 420 mg N, a difference of 156 mg N from the earlier estimate. This net change includes whatever water was removed, chemistry changed and input occurred. It is not automatically 156 mg of nitrogen transformed by bacteria. The exercise illustrates how easily an inventory comparison becomes a false process measurement when the boundary conditions are omitted.

FAQ: Can a single ‘normal’ TAN number clear a hatchery for the day?

No. A single TAN observation is one chemical measurement at one time. It may be useful, but pH, temperature, salinity, changes during the day, stage-specific response and sampling quality also matter. A valid reading does not cover unmeasured hours. Whether the measurement is adequate for an operational decision depends on a documented monitoring programme and a locally appropriate biological interpretation, not on the word ‘normal’ alone.

The same caution applies to dissolved oxygen. A noon reading cannot reconstruct an unlogged overnight minimum. Translate an observation into the time and place it represents, and separately name what was not measured. Where the consequence of a missed event is material, the monitoring design must be reviewed by those responsible for animal care and system operation.

FAQ: Does a fall in TAN prove that a biofilter is working?

Not by itself. Dilution, changes in feed, altered excretion, sampling variability and other processes can reduce TAN. To support a specific claim about nitrification, seek appropriate process evidence, including nitrite and nitrate measurements with consistent nitrogen units, flow and water-exchange accounting, and observations before and after a controlled change. Even these may not uniquely prove a process without additional design.

Use a statement proportionate to the evidence: ‘TAN declined over this interval’ is observed; ‘nitrification caused the decline’ is a hypothesis until supported. The distinction matters when a treatment decision or financial claim depends on the presumed mechanism. Keep competing explanations visible until data can discriminate between them.

FAQ: Can a reported ammonia percentage be used without its method?

Not reliably. A percentage could mean un-ionised fraction, analytical recovery, percentage change from baseline or even a component of an instrument’s internal calculation. For ammonia speciation, the fraction should be tied to the TAN basis, water temperature, pH, salinity or ionic strength assumptions and the selected equilibrium model. Without those elements, the reported percentage is not reproducible.

Ask for the underlying parameters or calculation note rather than reverse-engineering a plausible fraction from unrelated values. If data cannot be recovered, report the percentage as an unverified supplier estimate and avoid placing it in a decision-critical comparison. This is a translation of evidence quality as well as of chemical language.

FAQ: What is the best way to hand the log to another language team?

Supply a bilingual or multilingual field dictionary defining original column headers, units, reporting bases, instruments, specimen or tank identifiers, time convention and missing-value codes. Keep chemical formulae unchanged and explain them in full sentences. Mark measured versus model-derived fields, and attach enough source documentation for another reader to reproduce the conversion. A clear glossary does not replace training in sampling, but it prevents avoidable changes in what the data claim.

When translating a narrative conclusion, preserve uncertainty words such as ‘estimated’, ‘possible’, ‘below reporting limit’ and ‘at the sampled time’. Do not strengthen ‘may indicate’ into ‘proves’. A technically faithful translation can be easier to read without becoming more certain than the evidence. That is the difference between linguistic fluency and defensible scientific interpretation.