The number is an assay result before it is a growth claim
Marine bacterial leucine incorporation is a rate of labelled amino acid becoming associated with material recovered by an operational assay. In many studies the label is tritium and the recoverable material is protein-rich acid-insoluble material. The method reveals an aspect of heterotrophic prokaryotic protein synthesis. It does not directly weigh all newly grown bacterial carbon, measure whole-community respiration or count cell divisions. Those are additional inferential steps, not alternate names for the same observation.
A useful opening distinction is between incorporated leucine, inferred protein synthesis and converted carbon production. A rate such as 35 picomoles leucine per litre per hour describes a reaction associated with a measured water volume and incubation interval. Multiplying by a conversion factor can produce a model-derived biomass production rate, but that conversion needs evidence and assumptions. The analyst should show both values when teaching the result, rather than presenting the second as if the instrument measured it.
The lesson matters because a perfectly calculated incorporation rate may still be an uncertain estimate of biomass production. A report can have precise radioactive counts, small replicate scatter and appropriate blanks while its carbon conversion factor remains poorly constrained. Separate analytical precision from ecological interpretation. A reader can then recognise which part of a result can be defended by direct measurement and which part is a scenario-sensitive estimate.
Define the organisms and the reaction
Heterotrophic prokaryotes include bacteria and archaea, but not every member of a mixed community will take up the supplied substrate at the same rate. A leucine assay preferentially records the activity of organisms participating in labelled leucine uptake and incorporation during the incubation. It is therefore an index of a physiological process in the sampled community, not a complete census of every microbial pathway operating at the station.
Leucine is an amino acid used in protein. A cell can obtain it from the water, synthesise it internally, respire it or transform its carbon skeleton. The label reaching the acid-insoluble fraction is interpreted through that chemistry. A reader should never replace ‘protein-associated tracer’ with ‘number of bacteria’ unless there is an independent calibrated pathway from the measured activity to cell counts.
When species composition changes along a transect, the fraction of cells using leucine and the relationship between incorporation and actual biomass production can change. A comparison between warm surface water and dark deeper water therefore needs more than matched sample volume. Consider carbon substrate availability, growth state, community composition and metabolic pathways before explaining the difference as an exact change in bacterial population size.
Follow one tracer through the measurement chain
The operational sequence begins with collection of seawater, addition of known labelled leucine, incubation for a recorded duration, termination of metabolism, extraction or precipitation, separation of retained material and measurement of radioactivity. Processing blanks and a specific activity calculation turn counts into an incorporation estimate. Each link can introduce uncertainty. A concentration written in the ship log is not yet a synthesis rate until time, sample volume and radiochemical accounting have been applied.
The incubation is a deliberate perturbation: added leucine changes the availability of one substrate. That may reduce the need for cells to synthesise their own leucine, but at some concentrations it can also alter uptake kinetics or metabolism. Avoid calling the incubation an untouched picture of the natural sea. It is a controlled test that attempts to approximate a natural process while making a tracer measurable.
An audit-ready table should preserve sample identifier, depth, incubation duration, added concentration, nominal and verified specific activity, live and killed counts, recovery and counting efficiency when relevant. It should also record whether rates were reported as moles leucine incorporated or carbon production. Without this provenance, a later analyst may compute a mathematically correct value from the wrong bottle, wrong duration or wrong tracer dilution.
Specific activity converts counts to substance
Specific activity expresses radioactivity per amount of labelled compound. If a recovered sample contains a given radioactive signal, the amount of tracer inferred depends on the measured specific activity and the counting corrections. The unit might be becquerels per nanomole, disintegrations per minute per nanomole, or a supplier’s curies per millimole. Those are related units of activity, not interchangeable numerical labels.
Imagine a teaching dataset with 1,200 corrected disintegrations per minute and a working specific activity of 400 disintegrations per minute per picomole. The corresponding amount is 1,200 divided by 400, or 3 picomoles of tracer equivalents. If the source value was in becquerels, converting only the numerator and not the specific activity would introduce a factor-of-sixty error because one becquerel equals one disintegration per second.
The measurement must distinguish instrument counts per minute from disintegrations per minute. Counting efficiency, quench correction and background can affect that conversion. A live scintillation count of 1,500 counts per minute is not automatically 1,500 disintegrations per minute. Report which correction has already been applied so a second person does not silently correct it again or fail to correct it at all.
Blank subtraction is necessary but not magic
Killed controls and procedural blanks estimate non-biological retention, adsorption, residual solution and instrument background. Their role is to identify the amount of measured signal not credibly associated with active incorporation during the test. The subtraction must be performed in compatible units and at the same stage of processing. A dead control obtained with a different incubation or filtration sequence may not represent the blank process for the live bottle.
Suppose three live aliquots give 1,510, 1,475 and 1,525 corrected counts per minute, and matched killed aliquots average 175 counts per minute. The live mean is 1,503.3, yielding 1,328.3 background-adjusted counts per minute. The mean of the uncorrected live signals would overstate net incorporation by about 13.2 percent relative to the adjusted net signal. This percentage is an arithmetic teaching example, not a claim about usual ocean blank sizes.
A near-zero blank-corrected result should not be forced into a positive ecological story. If blanks are variable enough that the difference between live and killed samples is unresolved, a numerical rate can still be computed but may not be reliably distinguished from zero. Say what the detection criterion is, preserve negative noise-corrected estimates for auditing when appropriate and follow the method’s reporting protocol.
Time makes an amount into a rate
The measured tracer amount belongs to a particular incubation interval. A rate requires division by elapsed time after the blank and specific-activity corrections. Record actual start and stop times, not just the intended interval. Incubations at sea can be delayed by sample processing, temperature control or deck operations, and small timing differences matter when turnover is rapid.
If one millilitre of water incorporates 1.8 picomoles over forty-five minutes, divide by 0.001 litre and by 0.75 hour to obtain 2,400 picomoles per litre per hour. This is a deliberately constructed example chosen to demonstrate conversion of both a volume and a time denominator. Reporting 1.8 picomoles per litre per hour would lose a factor of about 1,333.
A time course should check whether incorporation is approximately linear over the selected window. If the cumulative label doubles between fifteen and thirty minutes but rises little between thirty and sixty, a single sixty-minute average may hide substrate depletion, saturation, changing community behaviour or procedural difficulty. The scientist should describe the interval actually tested rather than assume that every incubation duration measures an identical instantaneous rate.
Volume belongs to the seawater aliquot, not the counter vial
A sample may be incubated in one container and finally transferred into scintillation cocktail in another. The denominator for water-column volumetric incorporation normally comes from the original seawater aliquot analysed, not the larger vial volume after chemical addition. Reagent volume is important for quench and recovery, but adding it to the seawater denominator without methodological justification dilutes the reported biological rate incorrectly.
Imagine a two-millilitre seawater aliquot placed in a vial containing eighteen millilitres of counting mixture. If its tracer-equivalent incorporation is four picomoles per hour, its seawater-normalised rate is 4 / 0.002 = 2,000 picomoles per litre per hour. Dividing instead by the full twenty millilitres would report only 200. Both calculations have valid arithmetic; only the first uses the intended physical sampling denominator.
Some protocols use centrifugation and other protocols use filtration, with different aliquot handling. A plain-language explanation can say ‘per litre of original seawater tested’ and reserve reagent-volume details for the methods. This is a strong cross-language safeguard: it prevents a general term such as ‘sample volume’ from being interpreted as the final liquid volume inside the counting device.
Distinguish uptake, incorporation and growth
Uptake is entry of leucine or its label into an organism or associated material. Incorporation is recovery of label in the assay’s target fraction after processing. Growth is formation of new biomass, and it can involve many compounds and metabolic processes beyond leucine. These terms are related but cannot be substituted without supporting evidence. A cell may take up leucine and metabolise it without retaining its labelled atoms as newly produced protein.
An apparent incorporation rate can rise when uptake machinery becomes more active, when the fraction incorporated into protein changes, or when the community grows faster. A single tracer experiment does not automatically identify which mechanism dominates. This distinction is especially important when contrasting nutrient-rich coastal bottles with oligotrophic ocean samples, because carbon and amino acid pathways may differ.
Good technical translation uses staged language: ‘labelled leucine was incorporated into the recovered protein fraction at the measured rate’ followed by ‘under the stated conversion assumptions, this corresponds to an estimated prokaryotic production rate.’ The second sentence is conditional and informative. It does not undermine the usefulness of the first observation; it protects the reader from giving a derived quantity the certainty of a direct assay.
External isotope dilution changes interpretation
Unlabelled leucine already present in seawater mixes with added tracer. If the calculation treats the labelled supplement as though it were the only leucine available, the inferred tracer-specific conversion may be biased. Some experiments add sufficient labelled or unlabelled leucine to make ambient contributions comparatively small, while others characterise uptake kinetics across concentrations. Neither approach is assumption-free.
Consider a simplified label fraction: 2 units of labelled leucine mixed with 8 units of otherwise equivalent unlabelled leucine gives a tracer fraction of 0.20 before additional losses. If 1 unit of labelled leucine is incorporated, naively equating that to the entire leucine flux misses the unlabeled contribution. A simple fivefold scaling would apply only if pools mixed identically and the measured intracellular behaviour respected the model.
The distinction between extracellular dilution and intracellular synthesis is critical. A saturation curve may help address the first but does not prove the second absent. Label mixing in ambient water, de novo intracellular leucine synthesis and recycling after uptake are different mechanisms. State which factor is measured, inferred, bounded or ignored; one blanket phrase ‘isotope dilution corrected’ can conceal several incompatible corrections.
A high tracer addition can disturb the rate
Adding enough substrate to suppress isotope dilution can create an unnatural nutrient pulse. It may stimulate cells limited by amino acid availability or change competition for transporters. The result can become a potential rate under the added conditions rather than a pure measure of ambient in situ uptake. That is not automatically a failed experiment, but it changes the claim the result can support.
Imagine a concentration-response study measuring 12, 20, 32 and 36 arbitrary incorporation units as added leucine rises through four steps. The flattening from 32 to 36 suggests approaching saturation under that assay, but the plateau does not prove the whole community would behave at that rate in untreated seawater. No natural leucine concentration or growth rate can be reconstructed from that curve without a suitable kinetic model.
Report the actual added concentration, not simply ‘saturating amount’. Protocols differ across waters and investigative objectives. If the assay deliberately estimates maximum biosynthetic potential, say so. If it aims to approximate ambient production, justify how addition and incubation conditions affect representativeness. The strongest explanation is precise about the experiment’s intervention and modest about the ecology inferred.
Intracellular isotope dilution is a separate problem
Microbes can synthesise their own unlabelled leucine while taking up a labelled supplement. Inside the cell the incoming tracer may share a precursor pool with endogenously synthesised molecules. The label fraction in newly formed protein then depends on the intracellular pool, not merely on the proportions mixed in the seawater bottle. A calculation based only on external specific activity can miss this source of dilution.
Suppose the extracellular labelled fraction is 0.8 while the effective intracellular fraction reaching new protein is 0.5. The ratio 0.8 / 0.5 = 1.6 illustrates how a naive tracer-to-total incorporation conversion could diverge even with stable external concentrations. This example does not prescribe a universal correction: cellular leucine synthesis and mixing depend on physiology, conditions and assay design.
A reader should look for supporting saturation tests, comparisons with growth, isotope dilution experiments or sensitivity analyses. In the absence of those observations, do not select a correction factor because it makes the carbon budget look balanced. Publish the method’s reported rate, name assumptions and show plausible alternative outcomes when justified. Scientific honesty sometimes means leaving one conversion underdetermined.
Chemical recovery can distort the apparent incorporation
Precipitation, rinsing, centrifugation and filtration are operational steps that decide which labelled material reaches the counter. Loss of protein-rich material can depress recovery; retained free tracer or non-target molecules can inflate the apparent target signal. The assay measures what the method collects, so sample handling is part of the definition of the measurement and not merely laboratory housekeeping.
Consider a hypothetical extraction in which an internal check indicates eighty percent recovery of a relevant standard. Dividing a net tracer signal by 0.80 would yield a 1.25-fold correction if the check legitimately applies to the sample and analysis. But a standard’s recovery may not reproduce the behaviour of mixed microbial cells. Automatic correction without validating equivalence can replace one bias with another.
When reading different studies, record their precipitation temperature, extraction chemistry and handling method as reported, without assuming procedural variants have identical retention. Include method blanks and replicate processing checks. If an apparently dramatic difference tracks a protocol change rather than a water-mass boundary, it may be a laboratory effect rather than a true ecological gradient.
Counting uncertainty is not a biological error bar
Radiometric counting has a stochastic component: fewer recorded decays usually produce larger relative counting uncertainty. Sample preparation, quench adjustment and background subtraction contribute further analytical variation. Ecological variability across bottles or stations is a different source of spread. Mixing these into one unexplained error bar makes it hard to decide whether an interval is about detector precision or variability in seawater.
For a simplified Poisson example with 10,000 independent recorded counts, the square-root counting standard deviation is approximately 100 counts, or one percent of the observed count. With 100 counts it is roughly 10, or ten percent. These estimates describe an ideal counting component, not the total uncertainty of a leucine incorporation assay, which can be dominated by blank variation or conversion-factor ambiguity.
A graph caption should tell readers whether whiskers show replicate standard deviation, instrument uncertainty, confidence intervals or an assumed range in conversion factors. A tiny analytical error bar on a carbon-production figure does not make the biochemical conversion certain. Precision of the measured signal and validity of the ecological model must be evaluated separately.
Killed controls should be matched to the real workflow
A killed control cannot answer every contamination question if it is handled differently from live samples. If it receives tracer later, incubates at another temperature or remains exposed to reagents for a different duration, the control may represent a different physical process. The correct comparison is between a live aliquot and a control whose non-biological treatment is meaningfully matched.
Suppose a live sample has a background-adjusted signal of 900 units with matched killed control signal 150, while a separate unmatched control reads only 40. Subtracting 40 would produce 860 units; subtracting 150 would produce 750. The difference of 110 units is not an ecological finding. It is a warning that control definition can materially change the final calculation.
Documentation should identify how killing or metabolic arrest was performed, how quickly it occurred and when reagents were added. Such details are essential to audit, though they do not belong in every introductory sentence. Readers need the interpretation: background correction is only as relevant as the control’s relationship to the active experiment.
The conversion factor is an ecological model
A carbon conversion factor takes measured leucine incorporation into an estimate of prokaryotic carbon production. The factor embeds assumptions about protein composition, leucine use, intracellular dilution, respiration and the relationship between protein synthesis and whole-cell biomass. It is therefore unlike a constant such as the number of seconds in an hour. Different factors can lead to different carbon-production estimates from identical raw rates.
Take a measured incorporation rate of 25 picomoles per litre per hour. For illustration only, a factor of 1.0 kilograms carbon per mole incorporated leucine gives 25 nanograms carbon per litre per hour. Using 2.0 kilograms carbon per mole gives 50 nanograms per litre per hour. The conversion 25 picomoles times one kilogram per mole equals 25 nanograms follows unit cancellation; the factor values here are hypothetical sensitivity settings.
Published empirical and theoretical factors should be attributed, not silently transferred between marine regions. A 2022 methodological analysis by Giering and colleagues emphasised that common theoretical factors can overestimate production in some environments. That does not mean every historic result is invalid. It means the factor deserves its own uncertainty statement, especially when carbon budgets rely strongly on its magnitude.
Check the dimensional arithmetic before discussing a carbon budget
A measured incorporation rate has dimensions moles leucine per water volume per time. A factor expressing carbon mass per mole leucine changes its numerator but does not remove the volume or time denominators. The result remains a volumetric production rate. It can be integrated vertically to obtain an areal rate only with a depth calculation, and further integrated in time only with an appropriate temporal representation.
Suppose a profile reports 40 nanograms carbon per litre per hour across a uniform ten-metre layer, explicitly for a teaching simplification. Since a cubic metre contains one thousand litres, the volumetric rate is 40 micrograms per cubic metre per hour. Multiplication by ten metres yields 400 micrograms per square metre per hour. Renaming 40 nanograms per litre per hour as 40 micrograms per square metre per hour without the layer thickness is dimensionally invalid.
When a figure is supplied in micrograms carbon per litre per day, converting to hourly units requires division by twenty-four, not multiplication. A rate can also be reported in milligrams carbon per cubic metre per day; 1 microgram per litre equals 1 milligram per cubic metre. A writer who reconstructs units before rounding prevents errors larger than most of the biological signal being discussed.
A vertical profile is more than a single surface bottle
Marine bacterial activity can vary with depth because organic substrate supply, temperature, pressure, oxygen and microbial communities change. A surface aliquot therefore does not establish whole-water-column production. Integrating a profile requires representative sampling depths and an explicit method between them. Assigning a single measurement to an unsampled thousand metres creates a spurious total.
For a worked trapezoid example, imagine carbon production at depths zero, ten and thirty metres of 60, 40 and 10 micrograms carbon per cubic metre per hour. Integrating the first interval gives (60 + 40)/2 times ten = 500 micrograms per square metre per hour. The second gives (40 + 10)/2 times twenty = 500. The combined thirty-metre estimate is 1,000 micrograms, or 1 milligram carbon per square metre per hour.
That result is an integral over the sampled depth range, not over the entire ocean, and the piecewise linear interpolation is only one approximation. A strong report records the layer boundaries, missing depths and whether subdaily variability justifies treating the profile as steady. Whole-water-column extrapolation should be labelled as model-dependent when deep samples are absent.
Short incubations do not guarantee no ecological change
Using a short incubation can limit bottle effects but does not eliminate them. Temperature drift, light exposure, oxygen changes and altered particle encounters may act quickly. A deep-water sample warmed during deck processing can behave differently from water at collection depth. The incubation record should state controlled temperature and other conditions relevant to the organism and process.
Imagine matched water incubated at in situ temperature and five degrees warmer, producing 18 and 30 rate units. The 67 percent difference is a hypothetical demonstration of sensitivity, not a universal microbial temperature relationship. It shows why processing temperature deserves attention before concluding that a site naturally has high protein synthesis.
The most useful comparison uses matched protocol conditions across stations and additional tests of incubation artefacts where practical. If conditions necessarily differ, the narrative must be conditional. Say ‘the assay measured a higher incorporation rate under the stated incubation conditions’ before advancing an ecological explanation. This preserves exactly what the experiment supports.
Light and particles complicate a bacterial-only label
Marine communities include microbes attached to particles and free-living cells, with different access to substrates and different recovery characteristics. Some organisms conventionally grouped as microbial heterotrophs may react differently to light or particle-associated organic matter. Size-fractionation and filtration can change the community included in an assay. The operational microbial fraction is therefore a key part of the denominator.
Suppose a sample is filtered through a size cutoff that removes aggregates and large particle-associated cells. The measured uptake in the filtrate may be lower than in an unfractionated sample even if the remaining free-living cells have unchanged physiology. It would be wrong to describe the difference as an inhibitory effect of the filter without considering removed biomass.
The analyst must state whether rates cover total, free-living or operationally separated assemblages, and whether fractionation happened before or after incubation. A translation into everyday language can call these ‘microorganisms retained in the analysed fraction’, then define the size boundary. This avoids suggesting a sharp biological taxonomy where only a filtration procedure was used.
Compare leucine with counts and other production methods carefully
Cell abundance, thymidine incorporation, respiration and leucine incorporation probe different aspects of microbial ecology. Cell counts describe population size at a moment; radiotracer incorporation describes an activity during an interval. A high standing stock can coexist with low growth, while a smaller population can have a high per-cell uptake rate. One measure cannot automatically stand in for the others.
If abundance is two billion cells per litre and incorporation is twenty picomoles leucine per litre per hour, a community-average ratio of 0.01 attomoles leucine per cell per hour, or ten zeptomoles per cell per hour, follows arithmetic. The ratio is a community-average normalisation and does not prove that every cell incorporated ten attomoles. Some cells may be inactive while others dominate the signal. Cell enumeration errors also affect the ratio.
A reader should avoid declaring a disagreement simply because one method gives a different quantity. Increases in oxygen consumption describe energy metabolism, not the same mass of new biomass as a leucine-derived estimate. Comparisons become informative when the investigators explicitly specify what each method estimates and when a conversion or growth-efficiency model connects them.
Production, respiration and growth efficiency are linked but separate
Bacterial growth efficiency expresses the fraction of organic carbon processed into new microbial biomass rather than respired, under a specified accounting. Leucine-derived carbon production is often one input into such budgets. Without a compatible respiration measurement and assumptions about substrate carbon, it does not determine growth efficiency. Similarly, a high carbon production rate does not alone show whether a community is net releasing or retaining carbon.
A stylised budget might have bacterial carbon production of 20 arbitrary carbon-mass units per day and respiration of 80, yielding a growth efficiency of 20/(20+80) = 0.20, or twenty percent, if the two terms form the intended comparable carbon-processing budget. Changing the assumed production factor changes the efficiency. Production and respiration must share units, time windows and sampled communities.
Make the dependency visible when presenting such an efficiency. A reader should see that leucine incorporation was measured, carbon production was inferred and efficiency was calculated from production plus a second measured or estimated flux. Combining them into the phrase ‘the bacteria converted twenty percent of the available carbon’ overstates what the data may establish without substrate supply measurements.
Small differences can vanish within factor uncertainty
Suppose two hypothetical stations have incorporation rates of 22 and 26 picomoles per litre per hour. The second is approximately eighteen percent higher. But if one unknown ecological factor could vary between stations, the derived carbon-production ordering may not be robust. Replicate scatter, blank correction and conversion uncertainty should be considered before claiming a statistically or mechanistically meaningful difference.
Using one shared conversion factor preserves the ranking of the two values mathematically; using different uncertain factors can reverse it. If station A uses 2.0 carbon units per leucine unit and station B uses 1.0, the derived rates are 44 and 26 carbon units despite the lower raw incorporation at A. The example illustrates a modelling sensitivity, not evidence that different factors ought to be assigned to those stations.
When reporting an environmental gradient, provide the directly compared incorporation rates and explain the rationale for shared or varying factors. Do not inflate a modest difference through unjustified precision. Readers benefit from an explicit statement about what would have to change for the conclusion to reverse.
The endpoint can be detection-limited
At low incorporation rates, sample radioactivity can approach procedural background. A laboratory may use counting time, replicate assays, control variability and reporting thresholds to decide what values are quantifiable. Detection is not merely a property of the instrument; it is influenced by sample volume, incubation duration, recovery and blanks. A result below quantification should not be translated into an exact zero.
Imagine a blank-adjusted estimate of 3 rate units with an estimated analytical uncertainty of plus or minus 5 units. It is not defensible to print 3.000 as though the third decimal were informative. Depending on the documented method, the value could be reported as below quantification, with the measured estimate retained in supplementary records. Avoid censoring a whole profile without identifying which points fall below thresholds.
Do not increase incubation duration solely to force a signal above threshold while ignoring possible nonlinearity. A more sensitive assay may require longer counting, greater sample volume or another validated approach. The goal is an honest measurement with a known detection limit, not a large number obtained by changing the biological process.
Biological replicates are not instrument repeats
Three counting reads from the same extracted vial test counter repeatability. Three separate seawater aliquots incubated and processed independently can reveal more of the assay’s combined variation. Water samples collected independently at nearby depths may capture actual patchiness. The word ‘replicate’ needs this design information; otherwise a narrow standard deviation can be misread as strong evidence of spatial uniformity.
Suppose three separate live incubations after blank correction yield 20, 25 and 30 rate units. Their mean is 25 and their sample standard deviation is 5. If the same vial is counted three times and returns 24.8, 25.0 and 25.2, that spread is only an instrument-level check. The two kinds of variance cannot be interchanged when plotting site-to-site ecological uncertainty.
A methods section should say whether it describes technical, incubation or sampling replicates. It should also explain how killed controls were paired. If the source uses standard error of the mean, distinguish it from standard deviation. Readers can only assess the strength of a marine carbon inference if they understand what was genuinely repeated.
Check radiochemical and chemical labels in translation
Superscript isotope notation carries physical meaning. Tritium-labelled leucine, written ³H-leucine, is not equivalent to stable ¹³C-leucine, and both differ from unlabelled leucine. Reporting activity in disintegrations per minute is not the same as reporting chemical concentration. Translation should preserve isotopes, units and reaction targets even when simplified prose accompanies them.
The term ‘incorporation’ should not become ‘absorption’ in a way that suggests all tracer taken up is retained in protein. ‘Potential bacterial production’ should not become ‘actual population increase’. ‘Bacterial carbon production’ should not become ‘ocean carbon absorption’. These changes may sound fluent but reassign the measured process. Treat the method’s vocabulary as part of the scientific evidence.
An effective explanatory sentence can pair the technical and everyday language: ‘The assay measures radiolabelled leucine retained in an operational protein fraction, used as an index of microbial protein synthesis.’ A second sentence can show the carbon conversion step. This produces a readable explanation without disguising the technical distinctions that make the measurement reproducible.
A complete worked reading from counts to rate
Build an auditable example using hypothetical values. A one-millilitre aliquot incubates for thirty minutes. The detector gives 3,000 counts per minute for the live sample and 600 for the killed control. Counting efficiency is 0.50, so net activity is (3,000 − 600)/0.50 = 4,800 disintegrations per minute. Assume the verified effective working specific activity is 800 disintegrations per minute per picomole incorporated tracer.
Dividing 4,800 by 800 yields 6 picomoles tracer-equivalent in the aliquot over half an hour. Dividing by 0.001 litre and 0.5 hour gives 12,000 picomoles per litre per hour, or 12 nanomoles per litre per hour. This unusually round teaching example exposes every unit conversion; it is not presented as a typical environmental observation. If no isotope-dilution correction is justified, label it an assay-equivalent rate under the stated assumptions.
Next, test each stage independently. Change counting efficiency from 0.50 to 0.40 while keeping actual instrument counts and the assumed specific activity fixed: the inferred rate rises by 25 percent. Double the incubation time without changing recovered signal: the estimated rate halves. Double the seawater aliquot volume with unchanged recovered signal: it halves again. These sensitivities reveal where an overlooked factor can quietly transform the conclusion.
A sensitivity interval is more honest than one false precision
Consider a net incorporation estimate of 30 picomoles per litre per hour. If an ecological conversion factor is genuinely uncertain across a twofold interval, the resulting carbon production should be represented as a corresponding interval or scenario, not as a single value with many decimal places. Such an interval can be wide even when count precision is high. This is not a weakness in the arithmetic; it identifies the true interpretive bottleneck.
For a teaching scenario with factors of 0.8 and 1.6 kilograms carbon per mole leucine, carbon production is 24 and 48 nanograms carbon per litre per hour. If a writer chooses the midpoint, the reader should still know the bounds and why the factor range was selected. These numbers are illustrative; they are not recommended factors for all marine environments.
Separate random analytical scatter, systematic assay assumptions and ecological factor uncertainty. Quadrature propagation is appropriate only for certain kinds of independent uncertainty and does not resolve unknown model bias. Do not use a small propagated standard error to conceal broad disagreement about how leucine uptake relates to whole-community biomass growth.
An independent transfer problem for a new sample
Try a different bottle without borrowing the worked answer. A two-millilitre aliquot incubates for forty minutes. After appropriate counting corrections, live material gives 1,800 disintegrations per minute and the matched killed control 300. The effective specific activity is 500 disintegrations per minute per picomole. Find net incorporated tracer, volumetric rate in picomoles per litre per hour and the direction of error if someone uses total scintillation-vial volume.
The corrected tracer signal is 1,500 disintegrations per minute; dividing by 500 gives 3 picomoles. The aliquot is 0.002 litre and incubation time is two-thirds of an hour, so the rate is 3/(0.002 × 2/3) = 2,250 picomoles per litre per hour. Using the larger counting-vial volume would make the seawater-normalised rate too low. The problem checks all three denominators independently.
Now interpret the answer: it is a protocol-defined incorporation estimate, not an observed biomass accumulation of 2,250 picomoles carbon. A correct translation preserves the leucine numerator and identifies whether an isotope-dilution or carbon-production conversion was actually measured. This reasoning is more transferable than memorising the arithmetic alone.
Independent transfer problem for station comparisons
Station A yields thirty picomoles leucine per litre per hour and station B twenty-five, each with the same analytical protocol. A reader claims that A has twenty percent more bacterial biomass. Check the numerical statement and the biological claim separately. The ratio thirty divided by twenty-five equals 1.20, so A has a twenty percent higher measured incorporation rate under the same operational conditions.
The conclusion about biomass is unsupported. The assay records a process rate, not a standing stock. Even a statement about twenty percent more carbon production requires the factor or its comparability. If the same factor is justified for both, the model-derived rates retain that ratio. If factors differ for biological reasons, the ranking itself may need reevaluation.
The repaired explanation is: ‘Incorporation was twenty percent higher in sample A; the implications for biomass production depend on the factor and the communities compared.’ This sentence lets the data speak clearly while preventing a unit change from masquerading as biological evidence.
Practical quality checks before publication
Start with the raw measurement identifier, depth, date and storage history. Confirm live and killed sample pairing, time interval, original seawater volume, counter background, quench correction and specific activity. Then check whether the result is a tracer-equivalent incorporation rate, a total leucine estimate under a dilution model or a carbon-production estimate using a named factor.
Next inspect replicate design, linearity through time, response to tracer concentration and any instrument or extraction recovery concerns. Look for implausible jumps associated with processing batches or changes in laboratory method. Check whether concentration units describe a water volume or mass and whether per-day rates have been converted consistently. A reader should be able to reconstruct at least one representative calculation from the reported inputs.
Finally ask whether the ecological sentence reaches beyond the assay. A difference between stations may reflect substrate availability, physiology, community composition or environmental conditions; it does not automatically diagnose a single cause. A well-written result distinguishes observation, calculation and plausible interpretation, enabling readers to revisit the conclusion when new conversion-factor evidence emerges.
Frequently asked questions and bottom line
Does leucine incorporation count bacterial cells? No; it estimates incorporation of tracer into an operational fraction. Does a low rate mean bacteria are absent? No; organisms may be present but inactive or use other pathways. Does a saturated tracer addition reproduce natural substrate supply? Not necessarily. Does a tiny instrument error bar establish an accurate carbon-production estimate? No; the ecological conversion may be far less certain than the radiometric count.
Can different studies be compared? Sometimes, if incubation conditions, fractionation, isotope accounting, denominators, blanks and conversion factors are made comparable. Can a rate be turned into an areal flux by changing the unit label? No; integrate over depth with a justified profile. Can carbon production be equated with carbon respiration? No; they are distinct terms. These distinctions matter more than the number of decimal places retained.
The durable reading rule is to follow the label from seawater bottle to tracer signal to incorporation rate and only then to inferred production. Explain what is directly measured and where assumptions enter. If the reader can reconstruct the units and identify the most influential correction, the scientific meaning survives translation into clear language.
Primary sources and connected learning
Sources and related reading: Kirchman, K’nees and Hodson (1985), leucine incorporation and protein synthesis; Giering et al. (2022), conversion-factor overestimation and cautions; Smith and Azam (1992), marine leucine microcentrifugation assay; eduKateSG marine phytoplankton flow cytometry explanation; eduKateSG seawater particulate carbon and nitrogen explanation; eduKateSG series index. All numerical worked examples above are fictional instructional calculations; none are presented as field observations.