Why two isotope numbers are not a food-chain diagram
Stable carbon and nitrogen isotope measurements can help researchers investigate the sources of organic matter and relative trophic positions in marine food webs. Yet a dot on a δ13C–δ15N plot does not reveal a species’ entire diet. Each coordinate comes from ratios measured in sampled tissue, compared with reference isotope standards. Interpreting those ratios requires knowledge of baseline organisms, diet-to-tissue differences, sample preparation, temporal integration and uncertainty.
This guide explains how to read marine food-web δ13C and δ15N data with correct units, baselines and cautious ecological conclusions. We will calculate hypothetical trophic positions, compare different baselines, diagnose lipid effects, and work through independent problems with answers. Every numerical example is invented for learning; none is presented as a field observation or a new ecological dataset.
When a researcher writes that a fish has δ15N of +12 per mil, the measurement is not twelve percent nitrogen and does not imply that twelve percent of its food came from a particular prey species. The isotope notation describes a small relative difference in a ratio of heavier to lighter isotopes. Understanding that language is the first protection against an attractive but incorrect food-web story.
Define an isotope ratio before interpreting it
Carbon has stable isotopes including 12C and 13C; nitrogen has 14N and 15N. Instruments assess ratios such as 13C/12C and 15N/14N and compare sample ratios with recognised reference scales. The delta convention expresses the difference from the standard as δ = (R_sample/R_standard − 1) × 1000, with values in parts per thousand, written ‰.
A δ13C value of −20‰ means the sample’s 13C/12C ratio is approximately two percent lower than the reference ratio, not that the sample contains negative carbon or that twenty percent is the heavier isotope. The delta value is a relative description of an isotope ratio. It must not be relabelled as isotope abundance, concentration, mass percentage or food fraction without an appropriate conversion and model.
The choice of isotope reference matters. Carbon isotope values commonly relate to the VPDB scale, while nitrogen isotope values are commonly referenced to atmospheric nitrogen. Laboratory standards and calibration practices are needed to place results on those scales. In a translated report, preserve the δ symbol, isotope superscripts, unit ‰ and reference scale wherever the method gives them.
What the carbon coordinate can indicate
Carbon isotope composition can help distinguish organic matter derived from sources whose carbon isotope baselines differ. In marine settings, primary producers, habitats and water-mass carbon conditions can generate distinct δ13C patterns. Consumers carry a tissue signature influenced by assimilated diet and subsequent biochemical processing. That makes carbon useful as a source clue, not an automatic exact-diet calculator.
A fish at −16‰ and another at −21‰ need not occupy different trophic levels. They may draw from different producer baselines, consume different mixtures, differ in lipid content or reflect different tissue turnover periods. A single carbon coordinate cannot locate a fish on a map or tell us which named prey species supplied its food. Hypotheses require environmental baselines and additional evidence.
Mixing models sometimes estimate contributions from multiple sources, but they require identifiable endmembers, appropriate trophic discrimination assumptions, elemental concentrations where relevant, and a justified statistical model. Two isotope values do not necessarily identify three or more unknown diet fractions uniquely. A neat triangle on a graph can conceal uncertainty and overlapping sources.
What the nitrogen coordinate can indicate
Nitrogen isotope composition often increases from prey to consumer because of isotopic fractionation associated with nutrition and metabolism. That pattern can help approximate relative trophic position when the base of the food web and the relevant discrimination factor are known. The separation between consumer and baseline is more interpretable than the consumer’s absolute δ15N alone.
The trophic discrimination factor is the average change in tissue isotope composition relative to diet for a defined species, tissue and nutritional setting. It is not the same as a universal ecological constant. Experiments demonstrate that diet quality, temperature, physiology and other factors can matter. If a paper assumes a particular Δ15N per trophic step, the calculation should state and justify that assumption.
A high δ15N can arise because a consumer eats higher in the food web, because it feeds on an isotopically enriched baseline, or because both processes occur. Without a local baseline, comparing absolute values across ecosystems may create a fictional ranking. Baseline correction is therefore a necessary scientific step, not merely a statistical convenience.
Baseline organisms are part of the measurement
A baseline is an organism or material used to characterise the isotope composition at a known or assumed position in the food web. Depending on the system, researchers may sample suitable primary consumers or producers. The key is that the baseline should represent the energy pathway and spatial or temporal conditions relevant to the consumers being compared.
If two regions have baseline δ15N values of +5‰ and +9‰, a consumer measuring +12‰ in both regions has different offsets from the respective bases: +7‰ in one and +3‰ in the other. Treating both consumers as having identical trophic position from their absolute values would erase important geographic or biogeochemical differences.
Baseline uncertainty can be spatial, seasonal and taxonomic. A single baseline specimen may not represent a complex estuary or a water column with different nutrient sources. Researchers should sample enough reference material to estimate variation, identify the trophic level assigned to the baseline, and explain how that choice affects results. The inference is only as transportable as its supporting baseline data.
Calculate trophic position with named inputs
A commonly used simplified relation is trophic position TP = λ + (δ15N_consumer − δ15N_baseline)/Δ15N, where λ is the trophic position assigned to the baseline organism and Δ15N is the assumed diet-to-consumer enrichment per trophic step. The calculation is transparent, but its inputs are empirical assumptions rather than numbers supplied by the isotope instrument itself.
For an invented example, take a primary-consumer baseline at trophic position 2 with δ15N of +6‰, and a fish at +12‰. If a study justifies Δ15N = 3‰ per step, TP = 2 + (12−6)/3 = 4. This is an estimated trophic position under those specified conditions, not direct observation that the fish ate exactly two sequential prey organisms.
Change the discrimination factor to 4‰ while holding the isotope measurements fixed, and the estimate becomes 2 + 6/4 = 3.5. The disagreement is not arithmetic error; it demonstrates model sensitivity. If the factor is uncertain, the reported trophic position should include that uncertainty or an explicit sensitivity analysis.
An alternative baseline changes the answer
Imagine the same consumer at +12‰ but a baseline at +8‰, again assigned trophic position 2. Under Δ15N = 3‰, the estimate becomes 2 + (12−8)/3, approximately 3.33. A two-per-mil baseline shift changed the interpretation by about two-thirds of a trophic step even though the measured consumer did not change.
That sensitivity explains why comparative studies should not transfer a reference baseline from a distant water body without justification. A value that looks similar numerically across two sampling locations may have different ecological meanings. Conversely, different consumer δ15N values can represent similar trophic positions when their baselines differ.
A rigorous caption specifies whether plotted values are raw tissue δ-values, baseline-subtracted values or derived trophic positions. These quantities have different units and assumptions. Do not use the same colour scale label for all three without explaining the transformation.
Lipids complicate carbon interpretation
Lipids are commonly depleted in the heavier carbon isotope relative to some other biochemical tissue fractions. Variation in lipid content can therefore change bulk tissue δ13C even if the consumer’s assimilated carbon sources did not change in the way the plot might suggest. This is particularly relevant when comparing fatty and lean tissues, feeding conditions or organisms with different energy stores.
Researchers may chemically extract lipids or apply mathematical normalisation, but those procedures can alter results and are not interchangeable. A correction derived for one species and tissue type may be biased elsewhere. Some extraction methods also affect nitrogen isotope values, so the handling of carbon and nitrogen subsamples matters. Preserve the actual preparation protocol rather than simply stating ‘isotopes were corrected’.
If a carbon shift disappears after a validated lipid adjustment, the interpretation differs from a shift supported by independent diet evidence. Both raw and adjusted values may be informative, provided the method explains their roles. A reader should never be forced to infer which numbers were measured and which were model-corrected.
The carbon-to-nitrogen ratio is another denominator
Many studies include tissue C:N ratios to describe composition or guide a lipid correction. The ratio may be expressed by mass or by amount of substance, and those are not numerically identical. A mass ratio divides grams of carbon by grams of nitrogen; a molar ratio divides moles of carbon by moles of nitrogen. Converting requires the relevant atomic masses.
For example, 36 grams of carbon and 14 grams of nitrogen give a mass ratio of 36/14 ≈ 2.57. Those masses represent about three moles of carbon and one mole of nitrogen, so the molar ratio is approximately 3.0. Reporting ‘C:N = 3.0’ without its basis is ambiguous. A lipid-correction model calibrated on one ratio definition should not be fed values from another definition.
The C:N ratio does not directly reveal prey identity. It is information about sample composition and can support quality control and specific corrections. Treat it as an input with a defined scale and denominator rather than a universal diet signature.
Tissue turnover and timing
An isotope measurement reflects materials incorporated into the sampled tissue over a biological period; it is not necessarily a snapshot of the most recent meal. Different tissues integrate different time scales depending on metabolic turnover, growth, temperature and species. A fish’s muscle, liver and other tissues may record different aspects of its diet history even when sampled on the same day.
This affects comparisons with water conditions, plankton samples or telemetry. A fish captured during a brief event may still carry tissue signatures dominated by earlier feeding. Connecting the isotope measurement to a calendar event requires appropriate turnover information or longitudinal data. The statement ‘ate this prey yesterday’ is far stronger than most single tissue measurements can support.
Tissue selection is therefore part of the evidence. Reports should state the tissue, preservation, extraction, preparation, instrument and collection period. Comparing different tissues as if they were identical may create false ecological differences.
Replication must match the food-web question
Multiple mass spectrometer injections of the same powder provide analytical repeatability, not independent animals. Several subsamples from one fish examine within-specimen variation. Different fish provide biological replication, subject to the sampling design. If animals share a school, feeding ground or environmental baseline, their observations may also be correlated at larger spatial scales.
Food-web comparisons often involve nesting: tissues within animals, animals within sites, sites within regions and dates within seasons. A statistical model or carefully defined descriptive summary should respect that structure. Large numbers of isotope measurements do not automatically imply an equally large independent biological sample.
If a paper reports twenty-five observations from five fish with five tissue replicates each, the number of fish remains five. Uncertainty about an ecosystem or population cannot be replaced by instrument-level precision.
Precision does not supply source separation
Laboratories can measure isotope differences precisely, but if prey sources have very similar δ13C and δ15N compositions, their mixture fractions may be poorly identifiable. A narrow analytical error bar cannot resolve overlapping ecological endmembers. This is a different uncertainty: the model lacks a clear chemical contrast among proposed sources.
The converse is possible too. Endmembers may be distinct in one season and overlap in another. Using archived source values without testing temporal stability can lead to unjustified diet estimates. Good studies collect relevant sources, document their distributions and show how model results change with alternate assumptions.
A translation that states ‘diet was 40% species A’ needs the model, inputs and uncertainty that make this number meaningful. It must not be inferred from the position of a point by visual interpolation unless a validated and appropriately constrained mixture calculation supports that conclusion.
Food-web baseline shifts can imitate trophic shifts
Nitrogen at the base of a food web reflects nutrient sources and transformations. If the baseline becomes isotopically enriched, consumers may also show higher δ15N without eating at a higher trophic level. A before-and-after comparison therefore needs contemporaneous baseline information, particularly across years with changing oceanography.
A plot that shows fish δ15N rising over a decade is a valid empirical observation if sampling and calibration are consistent. It does not prove the fish became more predatory. Species size structure, tissue selection, source nutrients and baseline shifts are plausible alternatives. The inference should be tested rather than selected for narrative effect.
This logic also applies when translating maps: a colour gradient of isotope values may show spatial baseline variability, not a map of trophic position. Label the mapped quantity and avoid replacing its scientific name with a stronger interpretation.
The danger of universal enrichment factors
Textbooks sometimes illustrate trophic enrichment with a convenient single number. Such a teaching constant can help explain a formula but should not automatically become the field value in every species. Experimental work on Pacific cod found differing nitrogen trophic discrimination factors under different diets, demonstrating that assumed enrichment can depend on nutritional context.
If a study lacks a local factor, it may use a literature value with a clearly described rationale and sensitivity analysis. That is different from claiming the factor is a biological law. A responsible report states the source of the assumed Δ15N and explores how much the result changes when plausible alternatives are used.
Exact trophic positions in a complex ecosystem are model-derived summaries, not perfect discrete ranks. An estimate of 3.7 does not mean an organism consumed precisely 0.7 of a prey level. It is an index conditioned on baseline and enrichment assumptions.
Mixing models and more sources than dimensions
With two stable isotope dimensions, a researcher may wish to estimate multiple unknown diet contributions. But mixture identification can become weak when sources overlap or too many fractions must be inferred. Statistical priors, source variability and trophic corrections influence the result. A computer output with decimal percentages is not necessarily a uniquely determined ecological truth.
The interpretation also depends on whether elemental concentrations differ among food sources. An isotope mixture is not always a simple arithmetic mixture of bulk diet mass fractions when the sources contribute different amounts of the element being measured. Concentration dependence may need to be addressed in the specified model.
Readers should ask whether alternative diets can produce similar consumer coordinates, whether the posterior intervals are broad, and whether all relevant sources were sampled. An honest conclusion may describe overlapping plausible contributions rather than forcing one winning prey category.
Write claims that preserve the evidential boundary
A careful sentence can state: ‘Consumer tissue δ15N was elevated relative to a sampled baseline, consistent with a higher estimated trophic position under the reported discrimination factor; the result remains sensitive to baseline and diet assumptions.’ It is less dramatic than naming the exact number of prey steps, yet it conveys the observable pattern and the calculation clearly.
For carbon, a defensible sentence may say: ‘The tissue δ13C shift is consistent with a changed mixture of assimilated carbon sources, but source overlap and lipid effects limit exact attribution.’ The words ‘consistent with’ identify compatibility without pretending uniqueness. Stronger conclusions are possible when experiments, additional markers or independent diet observations support them.
Translation should maintain mathematical signs, isotope superscripts and the per-mil symbol while preserving qualified language. Replacing ‘suggests’ with ‘proves’ or ‘baseline adjusted’ with ‘measured directly’ changes the scientific claim even if the numbers remain untouched.
Worked cases: the isotope ratio and the ecological claim
Case one. A student translates δ13C = −20‰ as ‘the tissue contains minus twenty percent carbon’. The sign and unit have been misunderstood. The quantity is the difference of an isotope ratio from a reference scale, expressed in parts per thousand. The negative sign means the sample ratio is lower than the standard ratio. It does not represent negative material, a carbon mass fraction, or the proportion of one food source in the diet.
Case two. A sample has a ratio R_sample that is 0.985 times R_standard. The delta value is (0.985−1)×1000 = −15‰. The calculation checks the sign and the scale. A report using −1.5% would represent the same ratio difference in percent form, but must not quietly label that number as −1.5‰. Keep the original delta notation when reporting isotope evidence.
Case three. A primary consumer baseline is assigned trophic position 2 and δ15N +6‰. A fish has +12‰, and the analysis assumes Δ15N = 3‰ per step. The calculated trophic position is 2+(12−6)/3 = 4. The result is conditional on the baseline and factor; it does not establish exactly which prey species the fish consumed or prove that the food web contained four discrete feeding events.
Case four. The same fish and baseline from the previous example are analysed with Δ15N = 4‰. Estimated trophic position becomes 2+(12−6)/4 = 3.5. Nothing about the measured fish changed. The difference is a sensitivity to an assumed diet-to-tissue discrimination factor. A transparent result would report the input rationale or a plausible range rather than quietly selecting whichever number supports a preferred ecological story.
Case five. A consumer has δ15N +12‰ in two estuaries, but the first baseline is +6‰ and the second is +8‰. Using λ=2 and Δ=3‰ gives trophic positions 4.0 and approximately 3.33, respectively. The identical consumer isotope value does not imply identical trophic ranking. Regional nitrogen sources and food-web baselines can alter the biological interpretation while the instrument measures the same absolute ratio.
Case six. Two consumers have δ15N values +11‰ and +13‰, but their baselines are +5‰ and +7‰. Both have a baseline offset of +6‰ and, under the same λ and Δ, the same estimated trophic position. A report ranking the second consumer as unquestionably feeding higher because its absolute value is two per mil greater would ignore the baseline difference and misrepresent the calculation.
Case seven. The isotope instrument reports δ13C and δ15N to two decimal places for three injections of one tissue powder. The values agree closely. That supports analytical repeatability; it does not make three independent fish. A population estimate needs biological sampling, and one fish’s isotopic identity is not an independent observation each time its powder is placed in the instrument.
Case eight. Five fish contribute four muscle subsamples each. The report calls twenty independent animals the sample. Correct it to five fish and twenty analytical or within-individual subsamples. The appropriate uncertainty calculation must reflect clustering by fish. Repeated subsamples can help assess preparation and tissue heterogeneity, but they do not replace independently sampled individuals in population-level inference.
Case nine. A carbon isotope plot shifts toward more negative δ13C after a period of poor feeding. The writer immediately concludes the fish switched to a different prey species. Changes in body lipid composition may affect bulk tissue carbon values. Check lipid extraction, C:N data, tissue composition, and relevant source measurements before claiming a dietary shift. A changed plot position is an observation; the prey switch is a model requiring evidence.
Case ten. A laboratory applies an empirical lipid normalisation developed for one species to a very different marine organism. The corrected δ13C values look plausible, but appearance is not validation. Different tissue composition and physiology can change the correction relationship. The safer analysis reports raw values, checks calibration suitability and either validates a species-relevant approach or retains the uncertainty as an explicit limitation.
Case eleven. A tissue contains 36 grams of carbon and 14 grams of nitrogen in a hypothetical sample composition exercise. The C:N mass ratio is 36/14 ≈ 2.57. With approximate atomic masses, that corresponds to three moles of carbon and one mole of nitrogen, giving a molar ratio near 3.0. The two numbers are not interchangeable inputs to a correction model unless the expected denominator is identified.
Case twelve. A graph gives consumer δ15N +10‰ and baseline +4‰. A writer divides ten by four and calls 2.5 the trophic position. That operation has no basis in the stated model. With λ=2 and Δ=3‰, the estimate is 2+(10−4)/3 = 4. The baseline enters as a subtraction, the difference is divided by a discrimination factor, and the baseline’s trophic position is added.
Case thirteen. Source A has δ13C −18‰ and source B has −22‰, while a consumer reads −20‰. Under an idealised two-source, equal-element-concentration linear approximation and zero discrimination, the midpoint is compatible with a half-and-half mixture. But real diets may require concentration dependence, discrimination correction and uncertainty. Do not announce an exact 50% diet unless the assumptions are justified and other potential sources excluded.
Case fourteen. The same consumer is plotted inside a triangle formed by three isotope source signatures, and a figure legend declares its diet to be uniquely identified. With multiple source fractions and uncertain endmembers, many combinations may fit. A statistical mixing model needs priors, source variability, discrimination factors and credible intervals. A geometric position alone is not an unambiguous food-item inventory.
Case fifteen. A fish’s muscle records an earlier feeding period, while a water sample was collected only on the day the fish was caught. Comparing those observations as perfectly synchronous can mislead. Tissue turnover and the baseline’s temporal variability determine what comparison is appropriate. The repair is to describe integration times and sampling dates, and avoid claiming that the tissue directly recorded the current day’s water chemistry.
Case sixteen. Liver and muscle samples from the same fish yield different isotope signatures. The analyst assumes one measurement must be wrong. Different tissue turnover rates, compositions and biochemical processing can produce legitimate differences. Investigate quality control, but do not force the values to agree. State the sampled tissue and its interpretive time horizon before comparing animals or feeding periods.
Case seventeen. The measured trophic discrimination factor in an experimental study differs between diets. A literature review replaces the separate values with a universal constant without comment. That simplification can change inferred trophic positions. A better summary identifies the experimental species, tissue and diet conditions, states the factor used in the application, and shows sensitivity to plausible alternatives where they materially affect the conclusion.
Case eighteen. A long-term fish dataset shows δ15N rising year after year, and the headline claims the fish became increasingly predatory. The producer and primary-consumer baselines may also have shifted through changing nitrogen sources. Compare contemporaneous baselines and species size structure before attributing the trend to feeding level. A relative nitrogen isotope pattern is not a unique record of trophic behaviour.
Case nineteen. A mixing model returns a posterior mean of 42% from prey A and an interval spanning 15% to 70%. Reporting ‘exactly 42% of the diet was prey A’ conceals the large uncertainty. The defensible statement identifies it as an estimated contribution under the model and reports the interval, endmember assumptions and sampling limitations. Model numbers are not direct counts of swallowed prey.
Case twenty. Two prey groups have nearly identical isotope coordinates, yet the model is asked to distinguish them separately. The available tracers provide weak source separation. Adding decimal places or a more colourful figure cannot create identifying information absent from the data. Combining indistinguishable categories, adding independent markers or collecting more discriminating source observations may be more honest than reporting unstable individual percentages.
Case twenty-one. A plot is labelled δ15N in percent rather than per mil, but the numeric values come unchanged from an isotope laboratory export. That introduces a tenfold scaling error relative to the ratio difference. Restore ‰ or perform a mathematically correct conversion with a transparent numerical change. Retain isotope superscripts and the declared reference scale so the measurement remains unambiguous in another language.
Case twenty-two. A high δ13C value is interpreted as proof the fish lived in a particular coastal bay. Carbon signatures reflect assimilated organic matter and may overlap among habitats or move through food webs. Without validated local sources, movement records or complementary tracers, the exact location claim is unsupported. A safer conclusion describes a carbon-source signature compatible with certain hypotheses, not a unique geographic coordinate.
Case twenty-three. A statistical analysis pools fish from distinct species and tissues, then compares their raw δ15N values as though their trophic discrimination and tissue turnover were identical. That assumption may fail. Stratify or model the relevant biological differences and explain the inferential target. A sophisticated plot cannot compensate for an inconsistent unit of comparison or an untested shared enrichment factor.
Case twenty-four. A baseline sample is taken in spring, while consumers are collected in autumn after major seasonal changes. The writer uses the spring value as an exact autumn reference. Assess seasonal isotope variability and collect appropriately timed baselines where possible. If only an older baseline is available, retain that limitation in trophic-position estimates. Baseline mismatch is a substantive source of uncertainty, not a minor citation problem.
Case twenty-five. A species contains both juveniles and large adults in the dataset. Their δ15N values differ, and an article attributes the difference entirely to a change in food-chain height. Growth, size, habitat use and baseline exposure may vary together. Separate age or size classes, examine diet evidence and document how individual characteristics affect the isotope interpretation. The difference is meaningful but not uniquely diagnostic of its proposed cause.
Case twenty-six. A report compares an untreated high-lipid tissue with a chemically extracted tissue from a different species, yet labels both values ‘raw δ13C’. That erases the analytical distinction. Preserve whether each sample was measured before or after extraction or normalisation. Where the treatment changes nitrogen isotope results, analyse appropriate subsamples rather than assuming a carbon correction leaves every other parameter unaffected.
Case twenty-seven. An article uses ‘enrichment’ without specifying which isotope and which comparison. A reader cannot tell whether the author means higher δ15N relative to prey, a rise in δ13C over time, or increased elemental nitrogen concentration. Rewrite it as a quantified isotope difference between identified tissues and references, with units and uncertainty. Precision in vocabulary can prevent an entire causal interpretation from becoming ambiguous.
Case twenty-eight. A field team has excellent isotope precision but no samples of the dominant primary producer feeding the system. The consumer values are still valid measurements. Exact diet allocation and trophic position, however, may be poorly constrained. The correct outcome is to report the observation and explicitly identify missing baseline coverage, not to invent an endmember that makes the mixing model produce clean fractions.
Independent transfer tasks with answers
Task A. A reference ratio comparison gives R_sample/R_standard = 1.007. Calculate δ = (1.007−1)×1000 = +7‰. The positive value means the sample ratio exceeds the standard ratio by seven parts per thousand. It says nothing directly about the tissue’s total nitrogen concentration. A translator should keep the plus sign and per-mil scale in the final expression.
Task B. A fish has δ15N +13‰, a primary-consumer baseline is +7‰ at λ=2, and Δ15N = 3‰. The estimated trophic position is 2+(13−7)/3 = 4. If the factor instead equals 4‰, the estimate is 3.5. This comparison is a sensitivity check, not a reason to average the factors without ecological justification.
Task C. Two sites have fish δ15N +10‰ and +12‰, and baselines +4‰ and +6‰. Each consumer is six per mil above its local baseline. Under the same λ and Δ, their estimated trophic positions match despite distinct raw isotope values. Explain why direct comparison of absolute δ15N across sites would have produced a misleading ranking.
Task D. A chart has prey-source clusters that overlap almost completely in both δ13C and δ15N. Can a two-isotope model uniquely identify their separate contributions? Usually not without additional informative constraints. The correct response is to report weak identifiability, model sensitivity and alternative source groupings, not to manufacture certainty by quoting highly precise posterior means.
Task E. A corrected carbon isotope value is published without the uncorrected value or the name of the lipid-normalisation method. What is missing? The preparation and correction protocol, its validation domain and the distinction between measured and model-adjusted values. The result may still be useful, but independent replication and comparison are compromised. An evidence-aware report identifies that methodological limitation.
Task F. A sample is labelled as muscle while the laboratory actually used liver, and the isotope number itself is correct. Is the ecological interpretation unchanged? Not necessarily. Tissues can have different turnover rates and biochemical composition. Correct the metadata and reassess any timing or diet comparisons that depended on the presumed tissue identity. A valid instrument number can support a misleading conclusion when the sample description is wrong.
Frequently asked questions
Does a higher δ15N always mean a higher trophic level? Not by itself. Baseline isotope signatures, tissue-specific discrimination and physiology can differ among individuals and ecosystems. A higher value can be consistent with higher feeding position, but credible comparison requires an appropriate reference and explicit assumptions. Absolute δ15N is a measurement, while trophic position is a derived estimate.
Can δ13C identify a specific prey species? Sometimes distinct, well-characterised source signatures and additional evidence can support a source hypothesis. But many prey overlap and diet-to-tissue processing changes values. A single carbon coordinate should not be translated into an exact prey name or percentage without appropriate modelling and uncertainty.
Are trophic discrimination factors universal constants? No. They are empirical quantities associated with species, tissue, diet quality and experimental context. A convenient published value may be useful for a sensitivity analysis, but its transfer to another system needs explanation. Preserve the factor and its provenance whenever a trophic position is reported.
Why can two fish in different places have the same δ15N but different trophic positions? Their baseline nitrogen isotope compositions may differ. Trophic inference uses the consumer’s offset from a relevant baseline, not solely the absolute ratio measured by the instrument. Different baseline assignments and enrichment factors can lead to distinct derived trophic positions.
What is the simplest reliable reporting habit? State what was measured, what was corrected, which baseline was used, which discrimination factor was assumed, and what the calculation does not uniquely establish. This small separation between observation and interpretation gives readers a way to check the claim rather than trust a decorative food-web diagram.
Connected eduKateSG reading
The fish otolith Sr:Ca and Ba:Ca habitat-history guide explores another chemical inference in living organisms; elemental otolith signatures are not equivalent to tissue food-web isotope signatures. The acoustic fish telemetry guide provides an independent example of why observation and inference must remain separate. Neither method automatically determines an animal’s diet from a pair of stable isotopes.
Primary studies and methodology
Source 1. NOAA Institutional Repository, experimental Pacific cod trophic discrimination factors under temperature and dietary treatments, https://repository.library.noaa.gov/view/noaa/56647 . The study reports different nitrogen discrimination factors across diets, illustrating why universal assumptions need testing.
Source 2. NOAA Institutional Repository, stable isotope variability in zooplankton across the eastern Bering Sea, https://repository.library.noaa.gov/view/noaa/65354 . The study illustrates spatial and environmental variation in carbon and nitrogen baselines.
Source 3. NOAA Institutional Repository, species identity and trophic markers in aquatic food webs, https://repository.library.noaa.gov/view/noaa/24467 . The study cautions against unexamined transfer among taxa and ecological markers.
Source 4. Mintenbeck and colleagues (2008), lipid treatment and model bias in carbon stable isotope ratios, https://doi.org/10.1111/j.1095-8649.2007.01754.x . The research provides methodological context for the limitations of lipid extraction and normalisation.