Why a feed conversion ratio can mislead
A fish-farm summary reports an FCR of 1.4. That small number appears to describe efficient feeding, but it cannot be evaluated responsibly until the reader knows what entered the numerator, which biomass entered the denominator, which production period was counted, whether mortalities and partial harvests were included, and whether natural food contributed to growth. Feed conversion ratio, abbreviated FCR, is fundamentally a relationship between a quantity of feed and a quantity of animal production. It is not a directly measured biological property of a feed pellet.
In aquaculture, the common operational calculation is feed mass divided by biomass gain or animal production mass over a defined interval. For example, 1,400 kg of feed associated with 1,000 kg of production gives an apparent FCR of 1.4 kg feed per kg fish or shrimp. The units cancel into a ratio only when both masses use compatible units. The comparison is meaningful only if accounting boundaries and moisture conventions are comparable. A farm reporting its best-looking figure is not necessarily reporting the same quantity as another farm.
This guide shows how to translate apparent FCR, growth, standing biomass, stocking, survival, mortality and feed records into honest calculations. All numeric cases are hypothetical learning examples, not reported farm performances or recommended feeding rates. A conversion figure can help diagnose records and pose questions, but it cannot substitute for species-specific feed programmes, qualified aquatic-animal health advice or sound environmental controls.
First principles: feed input and biomass output
Every production system has inputs, outputs and internal changes. Feed is an intentional input, but the feed delivered to a tank is not necessarily eaten. Some pellets escape, dissolve or settle, and some nutrients enter water or microbial pathways. Animals also eat naturally occurring food in ponds or biofloc systems, so the feed supplied by the farmer may not be the only energy source. Growth in wet body mass includes water, protein, fat and mineral, making wet-weight gain a different chemical quantity from feed dry matter.
The basic apparent FCR calculation asks how many kilograms of supplied feed correspond to a kilogram of observed biomass gain, under an agreed boundary. It is a production-accounting measure, not a proof that one kilogram of feed is chemically transformed into exactly one divided-by-FCR kilograms of flesh. Biomass can be lost through mortality, escape, harvest or sampling, and it can enter a pond through stocking or transfer. Each movement changes what the final standing biomass means.
To read a farm table, reconstruct the mass flow before dividing. Identify opening biomass, incoming stock, outgoing harvest, deaths, other removals, and closing biomass. Determine whether feed is measured as issued from storage, delivered through machinery, consumed according to an estimate, or purchased during a financial month. Those figures may differ substantially. The denominator must refer to production during the same biological period as the feed numerator, not merely the weight of animals still swimming at the end.
The most useful formula and its boundary
A simple interval apparent FCR is supplied feed mass divided by net biomass production credited to the interval. In a closed cohort with no transfers or harvest before the end, net standing biomass gain is final live biomass minus initial live biomass. With harvests or transfers, a more complete mass-accounting relationship for biological production can be written: closing live biomass + harvested biomass + other outgoing live biomass + biomass of losses, if counting growth embodied in losses, minus opening live biomass minus incoming live biomass. That final choice about mortalities and losses is a deliberate accounting convention, not something mathematics can decide for the reader.
Many farm-level apparent FCR reports use harvest production or net harvested output, while growth-trial FCRs often use weight gain among surviving experimental animals and may account for deaths under a separate convention. Always write which method is in force. If all fish were stocked as substantial juveniles, dividing feed by total harvested weight rather than gain can overstate apparent efficiency. If deaths occur, dividing feed by only final survivors’ gained mass can make a cohort appear less efficient for reasons that include production losses rather than feed physiology.
The label FCR without formula is incomplete. When comparing two farms or studies, ask whether the denominator is total harvested wet weight, net biomass gain, net production including dead fish, or something else. The correct editorial response is often to preserve “apparent farm FCR, calculated as…” instead of deleting the qualifier to make the text shorter. The formula is part of the claim.
Worked example: straightforward cohort gain
Imagine 10,000 fish were stocked at an average 50 g each, giving opening biomass 10,000 × 0.050 kg = 500 kg. After a period with no transfers or deaths, there are still 10,000 fish averaging 150 g, or 1,500 kg closing biomass. Net standing biomass gain is 1,500 minus 500 = 1,000 kg. If 1,450 kg of pellet feed was distributed in the same period, apparent FCR based on net gain is 1,450/1,000 = 1.45.
A mistaken calculation uses 1,450/1,500 = 0.967, dividing by final standing biomass rather than growth. It counts the original 500 kg of animals as if created by the new feed. The mistake may make the feeding programme look more efficient than it was. Another mistake uses 50 g and 150 g in the formula while feed is in kilograms; the numbers must be converted consistently.
A useful reporting sentence says: “During the defined cohort interval, net standing biomass increased from 500 to 1,500 kg; 1,450 kg of supplied feed gives an apparent FCR of 1.45 kg feed per kg net biomass gain.” It also names that survival was assumed complete and no transfers occurred. This level of detail lets readers reproduce the calculation without assuming it measures feed digestibility or nutritional quality.
Stocking, counting and estimated biomass
Opening biomass often comes from a count multiplied by a sample-based average individual weight. The count may include uncertainty from transport, grading, manual counting or automatic counters, while average weight may be estimated from a sample that overrepresents large or easily caught animals. A farm report that displays biomass to the nearest kilogram can conceal much larger uncertainty in the animal count and mean weight. This matters because net biomass gain is the difference of two estimates, and errors in either estimate flow into FCR.
For a simple example, 20,000 stocked juveniles averaging 30 g imply opening biomass of 600 kg. If the true count were 19,500, opening biomass at the same weight would be 585 kg. That 15 kg difference propagates into the calculated gain even if feed records are exact. The effect may be modest in a large growth interval but considerable in a short trial with limited gain. A confidence interval on an average body mass does not automatically include uncertainty in survival.
Sampling design should be described: number of animals weighed, sampling dates, whether weights were individual or pooled, and whether the stock was graded by size. A single “average 200 g” value may not represent a multimodal size distribution fairly if sampling favours one cage. Where a report aggregates many tanks, it should weight by the numbers or biomass represented, rather than averaging tank means without considering their population sizes. Reliable FCR begins with credible denominator estimates.
Feed delivered, feed eaten and feed purchased
Suppose a storekeeper records 2,000 kg of feed purchased in May. A feeder log records 1,750 kg issued to production tanks, while 100 kg remains in a day-bin and another 150 kg is held in sealed bags. The quantity purchased is not the quantity delivered to fish in May. A procurement-based monthly ratio would mix inventory timing with biological growth. A feed-delivered ratio should use material actually supplied to the relevant culture unit during the matching period, after accounting for returns and transfers under the farm’s method.
Automated feeders can log programmed amounts that differ from actual discharge because of bridging, empty hoppers, calibration errors or residual pellets. Hand feeding can produce record discrepancies when bags are counted rather than weighed. Some feed may be spilled before reaching the water or rejected by animals. Apparent FCR often counts supplied feed because actual consumption is hard to measure, but the term “consumed feed” must not be used if only delivery was recorded.
A robust audit reconciles opening feed inventory, purchases or receipts, outgoing transfers, closing inventory and waste. It then links the actual feed delivered to each tank over the same dates used for biomass. A feed batch number or pellet type may also be relevant if formulations changed. The reader should not infer nutritional superiority from a difference that could be explained by missing bags or inconsistent stock records.
Worked example: reconcile the feeder log
A hypothetical farm begins a month with 600 kg of a specified feed and receives 2,000 kg more. It transfers 250 kg to another site, discards 50 kg damaged feed and ends with 500 kg in stock. Mass balance indicates 600 + 2,000 − 250 − 50 − 500 = 1,800 kg remaining as feed potentially issued to the local system, assuming no unrecorded movements. If feeders report only 1,720 kg delivered, there is an 80 kg discrepancy to investigate, not an automatic correction to either register.
If the production biomass gain was 1,200 kg, using 1,720 kg gives apparent FCR 1.433, while using 1,800 kg gives 1.500. The difference is operationally meaningful and comes from record ownership, not animal biology. A good report presents the reconciled figure only after determining whether the missing 80 kg was carried over, transferred, spilled or mislogged. Treat unresolved differences as data-quality issues.
The arithmetic is intentionally simplified. Real farms may hold multiple feeds, water-added feed, open bags, feed returned to stock or automated systems with calibration uncertainty. The lesson is to keep the numerator traceable and to avoid treating a warehouse purchase as though it were automatically an animal intake. A seemingly precise feed ratio can be undermined by a modest inventory discrepancy.
Survival, mortality and the denominator problem
Mortality alters the meaning of biomass gain. Feed supplied to animals that later die is part of the farm’s feed expenditure, but their bodies may not appear in final harvested biomass. Some growth-trial definitions include estimated biomass of mortalities in production calculations; commercial measures may count only realised harvest. Neither convention should be silently substituted for the other. The biological question “How efficiently did survivors grow?” differs from the operational question “How much feed did the farm spend per kilogram of saleable product?”
If a tank begins with 10,000 fish and ends with 9,000, survival is 90% only if the count is reliable and no other removals or introductions occurred. Survival percentage is a count ratio, not a direct mass-loss percentage. If larger fish were disproportionately lost, their contribution to biomass can be greater than the count-based 10%. Mortality timing matters too: a death early in the period and one near harvest carry different histories of feed consumption and growth.
A useful report can present both survivor-gain apparent FCR and a mortality-adjusted production measure, with formulas clearly labelled, rather than selecting whichever appears most favourable. Where dead biomass is unknown or only a rough estimate, use an uncertainty statement. Treating a mortality-adjusted ratio as an achieved harvest ratio would exaggerate delivered production. The reader needs to know whether dead fish were counted as biological growth, economic loss or both in separate metrics.
Worked example: mortality shifts apparent FCR
A hypothetical cohort opens with 500 kg live biomass. During the cycle, 1,700 kg feed is delivered. It closes with 1,400 kg of survivors, and there were no transfers or partial harvests. The net standing-biomass gain is 900 kg, giving apparent FCR 1,700/900 ≈ 1.889. Suppose an audited mortality ledger estimates 100 kg of dead biomass accumulated during the period and a chosen biological-production convention credits that entire 100 kg in its numerator of produced biomass. The credited mass gain is 1,400 + 100 − 500 = 1,000 kg, giving 1.70 under this alternative method.
Neither figure is universally “the true FCR.” The first evaluates feed relative to the change in living stock. The second credits the biomass that was produced but subsequently lost, using the estimated body mass at death. The two metrics answer different questions. Without knowing the weight of dead animals, merely multiplying a mortality count by final survivor weight can overstate losses if many deaths occurred early.
A responsible published table would name the first “feed/net live standing-biomass gain” and the second “feed/estimated total biomass production including mortality biomass.” It would also provide survival and harvested output separately. The lesson is that a calculation must not conceal whether a loss was counted in the denominator.
Partial harvests and transfers
Commercial production often involves repeated grading, movement between units and partial harvests. A cage that opens with 800 kg and closes with 900 kg may still have produced 1,000 kg of harvestable fish during the period if substantial biomass left mid-cycle. Ignoring harvested biomass would make the denominator look artificially small and the FCR artificially high. Adding the same harvested animals twice would do the opposite. Each mass movement must be assigned to a date and destination.
In a simplified tank with opening biomass of 800 kg, no introductions, 600 kg harvested halfway through and closing live biomass of 900 kg, credited net production under a harvest-inclusive convention is 900 + 600 − 800 = 700 kg. If supplied feed was 1,050 kg, apparent FCR based on that production is 1.50. The final standing biomass alone grew by only 100 kg, but it is not the appropriate production denominator for a cycle that included a 600 kg harvest.
Transfers require particular care. If an incoming tank receives 300 kg of fish from another tank, the incoming biomass is an input, not new production. For a multi-tank farm, the internal transfer appears as an outflow from one unit and an inflow to another; a farm-wide consolidation must cancel it rather than count both sides as harvest. This is why lot identity and dates matter when farm systems exchange animals across units.
Wet mass, dry matter and moisture content
Fish biomass in farm FCRs is generally expressed as wet weight, while feeds can be dry pellets, moist feeds, fresh whole fish, pastes or mixtures. A wet feed weighing 100 kg may contain far less dry matter than a 100 kg bag of pellets. Directly comparing their as-fed apparent FCRs can mistake water content for nutritional efficiency. FAO guidance explicitly cautions readers to compare like with like and to distinguish apparent farm ratios from feed-only experimental results.
If dry pellets contain 90% dry matter, 1,000 kg as-fed contains 900 kg dry matter. If a moist feed contains 40% dry matter, 1,000 kg as-fed contains 400 kg dry matter. For a comparable 500 kg wet-weight gain, as-fed FCRs would both be 2.0 if 1,000 kg of either was issued; dry-matter feed inputs would instead be 900/500 = 1.8 and 400/500 = 0.8. Those are different mass bases, not evidence that the moist feed is nutritionally superior.
Even dry-matter conversion alone does not compare digestibility, energy concentration, amino-acid balance or wastage. The useful reporting discipline is to attach feed moisture, feed type and animal wet-weight basis to the ratio. If a study expresses feed consumption as dry matter, do not silently relabel the result as an ordinary commercial as-fed FCR. The measurement basis determines what can reasonably be compared.
Natural feeding and the word apparent
Pond fish can obtain nutrients from plankton, periphyton and other naturally produced organisms, while some biofloc systems offer microbial biomass as supplementary food. A low ratio of supplied feed to wet-weight gain may reflect a meaningful contribution from natural food, not simply superior efficiency of the supplied formula. That is why “apparent feed conversion ratio” is often the more honest term for commercial ponds. It represents the farm’s observed feed expenditure relative to growth under its conditions.
Consider a hypothetical pond supplied with 600 kg feed and recording 500 kg biomass gain. Apparent FCR is 1.20. Suppose a controlled research design plausibly attributes 150 kg of growth to natural food, leaving 350 kg of feed-associated gain under its model. Then a ratio framed as supplied feed/feed-attributable growth would be 600/350 ≈ 1.71, but this attribution is a separate biological model and cannot be recovered from a farm ledger alone. A casual calculation that subtracts an arbitrary “natural food bonus” manufactures precision.
The same farm might have different natural productivity in sunny and cloudy months or with different fertilisation and water exchange. Apparent FCR therefore combines feed practice and environment. It should not be promoted as a universal nutritional property of a product. The correct translation includes the context that made the number possible and recognises when natural food is unmeasured.
Feed cost per kilogram of gain
The lowest FCR is not automatically the lowest cost. Feed prices, ration composition, animal health, labour, electricity, production duration, survival and market value all matter. A farm may accept a slightly higher apparent FCR if the feed is less expensive and growth or survival remains satisfactory, or may pay more per kilogram for a formulation that achieves better production. Cost per kilogram of biomass gain is feed cost divided by gain, calculated on a consistent boundary, not simply FCR without prices.
Suppose Feed A costs $1.80 per kg and gives apparent FCR 1.30 under comparable conditions. Its feed expense per kilogram gain is 1.30 × $1.80 = $2.34. Feed B costs $1.25 per kg and yields FCR 1.65; its feed expense per kilogram gain is 1.65 × $1.25 = $2.0625, about $2.06. In this constructed example, the higher-FCR feed has lower feed cost per unit gain. Whether it is the better operating choice also depends on production time, water quality, survival and product specifications.
Avoid presenting price calculations without currency, period and feed basis. A budget prepared in one currency cannot be compared directly with another without an identified exchange rate and date. Feed cost should also be distinguished from full production cost: aeration, water treatment, labour, depreciation and losses may dominate some systems. An honest interpretation is often more useful than a simplistic ranking of ratios.
Specific growth rate and FCR measure different things
Specific growth rate, frequently abbreviated SGR, is often calculated from logarithms of individual or biomass weight over a time interval and expressed as percent per day. It describes relative growth pace under a stated convention. FCR describes feed input relative to mass gain. A faster-growing cohort can still have a higher apparent FCR, and a lower FCR can coincide with a longer production cycle. Neither quantity alone determines profit, welfare or sustainability.
For a hypothetical group with average weight increasing from 50 g to 100 g over 30 days, an individual-weight SGR is [ln(100) − ln(50)]/30 × 100, approximately 2.31% per day. That calculation is a weight-growth statistic. Without feed records and biomass accounting, FCR cannot be inferred from it. If fish were lost, a mean individual weight could rise while total standing biomass changed little.
A report saying “FCR increased because SGR increased” is a hypothesis, not a mathematical identity. Changes in maintenance metabolism, density, feed wastage, temperature, size and growth stage can alter both. Keep the numerator and denominator of each indicator visible and compare same species and comparable stages. In multilingual research reporting, preserve which logarithm and percentage convention were used before copying values into decision tables.
FCR over a month versus a complete cycle
A complete-cycle FCR aggregates feed and production across a defined cohort. A monthly or fortnightly ratio can be useful operationally, but short intervals are sensitive to sampling error, stocking changes and measurement lag. If estimated biomass gain during one week is close to zero, the calculated feed/gain ratio can become extremely large or even undefined. That does not necessarily mean the feed itself suddenly lost all nutritional value; the observed denominator may be too small or uncertain.
When combining intervals, do not take the simple arithmetic mean of interval FCRs. The correct aggregate based on identical conventions is total feed across intervals divided by total credited production across intervals. For example, suppose period one used 100 kg feed for 100 kg gain, FCR 1.0, and period two used 900 kg feed for 500 kg gain, FCR 1.8. The simple average 1.4 is not the cycle result. Total feed is 1,000 kg and total gain 600 kg, so aggregate FCR is about 1.667.
This is a ratio-aggregation problem with unequal denominators. It becomes more complex when movements and mortality accounting differ between periods. Preserve the raw feed and production totals in reports so that future readers can recompute the combined ratio. A dashboard that retains only the two FCR numbers may be mathematically insufficient for a reliable cycle summary.
Worked example: grade and move fish between tanks
Two tanks begin a period with 700 kg and 400 kg of fish respectively. During the interval, 200 kg live fish are moved from Tank A to Tank B. Tank A ends with 650 kg, while Tank B ends with 650 kg. Ignoring other movements and losses, Tank A’s credited production is 650 + 200 − 700 = 150 kg. Tank B’s credited production is 650 − 400 − 200 = 50 kg. Together the tanks produced 200 kg, matching combined closing biomass 1,300 minus opening biomass 1,100.
If 250 kg feed was delivered to Tank A and 100 kg to Tank B, apparent FCRs on these tank-specific credited gains are 250/150 ≈ 1.67 and 100/50 = 2.0. The farm aggregate is 350/200 = 1.75. A careless report that counts the outgoing 200 kg from A as harvest but fails to subtract the corresponding introduction to B would add 200 kg of fictitious production.
The demonstration assumes the transfer mass was measured, mortality negligible and all fish remained within the farm boundary. In practice, transfer handling can alter fish weight, water retention and stress, and a scale might report wet weights including water. Accounting should preserve both sides of a transfer under one linked identifier. The conclusion is not that one tank’s feed is better; each tank’s conditions, starting sizes and estimation uncertainty differ.
Survival and yield must be shown separately
A ratio that appears efficient can conceal severe mortality or poor total output. Imagine a hatchery that achieves a low apparent FCR among surviving animals but loses a substantial fraction of stocked juveniles. The biological performance of survivors and the economic efficiency of the whole cohort are different questions. Reporting FCR alone would not communicate the outcome. Survival, biomass yield, harvest size distribution and feed cost should accompany it.
Survival is commonly final surviving count divided by initial count, subject to introductions and removals. Yield is harvested biomass per relevant production unit, such as kilograms per pond, cubic metre, hectare or cycle. Production rate may involve biomass per unit area per year. The words yield and productivity must retain those denominators. A pond producing 5,000 kg over six months is not automatically twice as productive as one producing 5,000 kg over twelve months without considering resource use and physical area.
In longform explanation, distinguish an outcome from an efficiency ratio. Two farms with the same FCR can have different yields, exposure risks and financial performance. If a feed study publishes a high survival rate, check whether survival is per replicate, per production tank or weighted across different initial counts. An unweighted average of tank survival percentages may misrepresent the overall population when tank sizes differ.
Uncertainty in feed and biomass matters
An FCR estimate combines uncertain feed delivery and uncertain biomass gain. Counting bags may be highly accurate while actual pellet discharge is not, and body-weight sampling may carry error even when scales are calibrated. A denominator calculated by subtracting two large biomass estimates can have high relative uncertainty, especially over a short interval. Software displaying three decimal places cannot remove those measurement limitations.
Suppose 1,500 kg feed was delivered with an estimated possible discrepancy of 30 kg. Estimated biomass gain is 1,000 kg, with plausible alternative values from 900 to 1,100 kg because of sampling and survival uncertainty. Simple sensitivity bounds are 1,470/1,100 ≈ 1.34 and 1,530/900 = 1.70. This is not a statistical confidence interval, because the input ranges are not probability distributions. It does show why an apparent FCR quoted as 1.5000 may be misleading.
A good report records uncertainty sources, sampling frequencies, stock movements and whether feed totals were independently reconciled. When comparing small differences between treatments, the study design, replicate tanks, variance and statistical analysis matter more than a colourful bar chart. It is inappropriate to claim Feed A significantly improves FCR simply because its point estimate is 0.05 lower in two unreplicated tanks.
Performance comparisons require matched conditions
Feed A and Feed B might be tested in tanks with different temperatures, oxygen conditions, biomass densities, genetics, health histories and ration schedules. A lower FCR in one tank does not isolate feed formulation as the cause. Trials designed for causal comparison should assign replicates, account for tank-level dependence, use suitable controls and report growth, mortality, water quality and management consistently. Treating each fish within one tank as an independent experimental replicate can exaggerate statistical evidence if the feed treatment was assigned to the tank.
Commercial comparisons across farms have even more confounding factors. Apparent FCR is affected by natural food, water quality and production stage, not just pellet nutrition. To interpret a reported improvement, examine whether the before and after periods share comparable stock size, feed moisture, mortality convention, sampling design and cumulative exposure. If multiple changes occurred together, attribution needs stronger evidence than a simple trend.
In translating a study abstract, distinguish association from controlled treatment effect. “FCR was lower after a feeder change” does not mean “the feeder caused a reduction” unless design and analysis support that conclusion. Similarly, “the feed achieved 1.2” should identify whether the figure is mean, median, best replicate or combined aggregate. Words such as average, trial, treatment and statistically significant are technical claims with obligations.
The environmental boundary: feed not converted to harvest
Unused feed and waste-derived nutrients can create water-quality burdens, yet subtracting harvested animal biomass directly from feed mass does not give the mass of pollutants released. Feed contains carbon, nitrogen, phosphorus, water and minerals, while animal biomass has a different elemental composition. Some feed is respired into carbon dioxide; some nitrogen is excreted, incorporated into microbial biomass, retained in animals or removed in sludge and water. FCR is a convenient production indicator, not a complete nutrient budget.
A nitrogen balance requires feed quantity and composition, animal nitrogen retention, excretion, solid waste, outflow and other transformations. For example, 1,000 kg of feed at 30% crude protein contains roughly 300 kg protein on the stated basis, but conversion from crude protein to nitrogen depends on the analytical convention and cannot be equated with 300 kg elemental nitrogen. Even with a nitrogen estimate, nutrient release cannot be inferred solely from the FCR.
This is why environmental claims need measured or modelled processes and explicit assumptions. An FCR reduction might help lower feed use per kilogram produced in comparable systems, but it does not automatically prove less oxygen demand, less discharge or lower lifecycle climate impact. Reporting responsibly means stating which indicator was measured and which environmental outcome remains to be evaluated independently.
Feed storage, spoilage and quality control
A farm may report a deteriorating FCR when the underlying problem is feed handling. Poor storage, moisture damage, inappropriate particle size, pellet stability, feeding timing or equipment maintenance can increase wastage or change consumption. These possibilities should be investigated rather than asserted from FCR alone. A batch label and manufacturing date help trace a change, but do not prove causation without relevant comparisons.
Feeds delivered in different forms can have different moisture and nutrient concentrations. If an automatic feeder dispenses 50 kg of material but 10% remains clumped in the hopper, the issue is not just nutritional. A well-defined feed delivery measurement should identify whether the scale counted bag removal, hopper decrement or pellets actually entering water. A discrepancy may also indicate an unrecorded transfer or an instrument calibration issue.
In the diagnostic report, show records alongside the ratio: feed batch, ration changes, storage conditions, temperature, oxygen and animal health observations. Do not promise that changing a pellet will correct every high FCR. The value of the ratio is in prompting a controlled investigation of a production system that includes feed, animals, water, equipment and people.
Independent transfer: final biomass is not growth
A tank starts with 300 kg fish, ends with 900 kg and receives 900 kg feed. The net standing-biomass gain is 600 kg, so apparent FCR is 900/600 = 1.50. A trainee divides 900 kg feed by the closing 900 kg biomass and reports 1.0. The error is that initial fish biomass was incorrectly counted as new production. The repair is to state the chosen denominator and its source records.
If the tank had received an additional 100 kg live fish during the cycle, credited net production under a movement-inclusive method would be 900 − 300 − 100 = 500 kg, and the ratio would be 900/500 = 1.8. The example shows why a livestock inventory is necessary. A field label that merely says “stock weight” does not tell whether it represents an opening balance, an incoming transfer or final output.
Independent transfer: partial harvest changes the answer
A cohort begins with 400 kg live fish, harvests 300 kg during a cycle and closes with 600 kg, with no new stock or mortality. Total credited production is 600 + 300 − 400 = 500 kg. If 750 kg feed was delivered, the harvest-inclusive apparent FCR is 750/500 = 1.5. Ignoring the 300 kg partial harvest gives feed divided by 200 kg standing-biomass gain, or 3.75, which misstates the period’s total production under the chosen convention.
This example does not automatically make 500 kg the denominator under every commercial definition: some reports count harvested output alone, and research designs may handle losses differently. The task is to attach the formula to the reported number. A ratio without the denominator convention cannot be compared fairly to another farm’s headline value.
Independent transfer: mortality counted differently
A hypothetical trial uses 1,200 kg feed, starts with 200 kg fish and ends with 800 kg survivors. Dead biomass recorded during the period totals an estimated 100 kg. Net live gain is 600 kg, producing feed/live-gain ratio 2.0. If a study explicitly credits dead biomass within a total-production convention, its denominator is 800 + 100 − 200 = 700 kg, giving about 1.71. These two figures answer different questions and cannot be treated as a single corrected result.
Your editorial task is to write both labels and state the source of dead-biomass estimates. The mortalities might have happened at different fish sizes, making a late estimate uncertain. The alternative convention does not erase the economic or welfare consequences of losses. It simply defines a different denominator for a different analysis.
Independent transfer: combine two intervals properly
In one interval a farm distributes 200 kg feed for 100 kg biomass gain, FCR 2.0. In a second interval it distributes 400 kg for 400 kg gain, FCR 1.0. The arithmetic mean of ratios is 1.5, but the aggregate cycle result is (200 + 400)/(100 + 400) = 600/500 = 1.2. Explain why: the second interval contributed four times as much credited gain and must carry its proper denominator weight.
The calculation assumes the two intervals use the same feed and production definitions and no unaccounted stock movements. If monthly records use different mortality conventions or feed-moisture bases, summing their totals is not sufficient without reconciliation. This is a transferable lesson for any ratio: combine underlying quantities, not the headline ratios, when their denominators differ.
Independent transfer: feed price versus conversion
Feed A costs 2.20 currency units per kg and achieves an apparent FCR of 1.2 in a comparable, controlled hypothetical example. Feed B costs 1.40 per kg and achieves 1.7. Feed expense per kilogram of biomass gain is 2.20 × 1.2 = 2.64 for A and 1.40 × 1.7 = 2.38 for B. Feed B is cheaper per kilogram gain on this limited measure, even though its FCR is higher.
It would still be wrong to pronounce B universally superior. Differences in growth time, survival, water quality, market size, energy and labour are absent from the constructed comparison. The correct translation says the calculated feed-cost component is lower under the stated inputs. It does not claim a complete economic optimum or a validated on-farm recommendation.
Independent transfer: dry matter versus as-fed feed
A farm uses 1,000 kg moist feed containing 50% dry matter to achieve 400 kg wet fish gain. Its as-fed apparent FCR is 1,000/400 = 2.5. Its dry-matter feed/gain ratio is 500/400 = 1.25. A second farm reports 1.4 using pellets on an as-fed basis with 90% dry matter. A reader cannot fairly conclude from 1.25 versus 1.4 that the first feed is more efficient because the measurement bases differ.
To compare, first normalise the feed quantity on a consistent moisture basis, then still investigate differences in nutrients, digestibility, species, size and environment. The exercise teaches how easy it is for apparently exact ratios to become false comparisons when one word—dry, wet or as-fed—is omitted.
Independent transfer: measurement uncertainty dominates the ranking
Two farms report FCRs of 1.48 and 1.51. The first farm estimates fish numbers from a small sample and has unverified feeder calibration; the second uses audited inventory, biomass sampling and mass-balanced delivery logs. The 0.03 difference should not be declared a reliable efficiency advantage without estimating uncertainty and checking that denominator definitions match. Three decimal places do not create experimental precision.
The correct analysis records the uncertainty sources and asks whether the apparent difference exceeds plausible measurement variation. If the figures come from one tank each, confounding by husbandry and biological variation is likely. Your task is to distinguish a descriptive difference from a causal claim. The answer is that the data show two reported point estimates, but not enough evidence to rank feed quality with confidence.
Independent transfer: compute production with internal transfers
A farm contains Tank X and Tank Y. Tank X opens with 500 kg, transfers 200 kg to Y and closes with 550 kg. Tank Y opens with 300 kg, receives 200 kg and closes with 700 kg. Tank X credited production is 550 + 200 − 500 = 250 kg. Tank Y credited production is 700 − 300 − 200 = 200 kg. The combined production is 450 kg, equal to combined closing 1,250 minus opening 800.
If total feed delivered was 675 kg, farm apparent FCR is 675/450 = 1.5. A report that includes the 200 kg transfer as harvested output from X and then also counts it as new gain in Y would double count biomass. Matching transfer IDs solves this ownership issue. The exercise reinforces that boundaries, not merely numerical arithmetic, determine the meaning of every production ratio.
Frequently asked questions
Does a low FCR always mean a feed is good? No. The result can be affected by natural food, unrecorded feed, species, stage, mortality, stock estimates and accounting. Is FCR always calculated as feed divided by harvested mass? No; many uses refer to net biomass gain, and conventions differ. Is lower always better? Within comparable, well-defined conditions it often means less supplied feed per unit credited gain, but cost, survival, speed and welfare still matter. Can a commercial FCR be less than one? It can under particular accounting and natural-feeding conditions; do not infer that mass was created from nothing.
Can two monthly FCRs be averaged to get a cycle FCR? Not reliably without their feed and gain denominators. Does mortality-adjusted production represent saleable harvest? No. Can feed purchase records replace delivered-feed measurements? Not without an inventory reconciliation. Does feed price per kilogram establish economic efficiency? Not alone. Should an FCR result be interpreted as a digestibility trial? No; digestion, nutrient retention and production are distinct measurements.
The strongest interpretation starts with a ledger and ends with a qualified sentence: for a specified cohort, period, feed mass basis, production boundary and loss convention, a documented amount of feed corresponded to a documented amount of credited biomass gain. Any explanation of the cause or value of that ratio must use additional evidence.
Primary sources and related reading
These numerical examples are hypothetical and deliberately use clearly defined denominators. Real farm FCRs require site-specific, traceable inventory, mortality, feed and water-quality records. The linked sources support the definitions and cautions, not any unobserved farm-performance claims.
FAO, Why Should I Feed? apparent feed conversion and moisture limits
FAO, Freshwater Fish Culture Management: Determining and Using FCR
eduKateSG, How to Translate Easily Series Index
eduKateSG, Seafood Traceability and Mass Balance