The nursery question after settlement
A nursery report can announce that abalone seed has entered grow-out even though it records only a transfer from settlement plates to baskets. That movement is an important operation, but it is not proof that every animal was recovered alive, has adapted to a new feed, or meets the next stage’s release criteria. To translate the report accurately, first identify the biological stage, the date of the count, the physical container, the counting method and the size grade. An early juvenile attached to a plate, a detached juvenile acclimating in a basket and a graded animal ready for a larger grow-out unit should not be collapsed into the same English label.
This guide concentrates on the nursery interval after successful benthic settlement. It does not repeat the separate problems of fertilisation, swimming-larva competency or the first decision that an attached animal has metamorphosed. Instead, it follows a traceable juvenile cohort through collection, transfer, housing area, algae transition, feeding records, measurement, mortality reconciliation and readiness for onward movement. The examples use hypothetical training data, not observations from a real hatchery. They show how to preserve denominators, units and uncertainty so that a reader can reconstruct what actually happened rather than accepting the optimistic wording of a production summary.
Definition lock: species, stage and cohort
Write down the species name before interpreting a nursery protocol. An account for tropical Haliotis asinina cannot be converted automatically into husbandry guidance for temperate Haliotis sorenseni or Haliotis discus hannai. Species differ in thermal performance, feeding response and growth. The Southeast Asian Fisheries Development Center describes an intermediate nursery transition in H. asinina from about 10–15 mm early juveniles toward 25–30 mm advanced juveniles. Those figures describe that particular culture pathway, not an international definition of juvenile or marketable size. The translator should carry the species and local protocol beside any size threshold so that a number is not detached from the conditions that give it meaning.
A cohort means animals connected by an explicitly defined origin, not simply animals currently sharing a tank. A nursery basket may hold individuals from one spawning, several settlement dates or an emergency consolidation. If the origins differ, the count must retain a parent cohort identifier and a history of each merge or split. The same animals must not appear as a new production batch merely because the staff renamed a receiving tank. A reliable record has a stable identifier, start and stop dates, source containers, transfer counts and destination identifiers. Without them, a reported survival percentage may be numerically reproducible yet biologically uninterpretable.
Separate the production events
A nursery entry normally contains at least five different events: an estimate of animals available for collection, a collection or detachment event, a count actually accepted into the nursery, a sequence of husbandry observations and a final verified outgoing count. Each event answers a different question. The count visible before a plate is lifted is a census estimate; the number placed in a receiving basket is a transfer count; the number still present weeks later is a stock estimate. Calling each of these a hatchery yield masks losses between events. An honest translation names the event before selecting any percentage and makes clear whether the number is measured or inferred.
Make a simple movement ledger before calculating anything: opening animals plus confirmed additions minus known deaths minus confirmed transfers out equals expected closing animals. Expected and counted closing stock can differ because of inaccessible hiding places, counting error, escape or an omitted movement entry. Label the difference as an inventory discrepancy, not automatic mortality. If a nursery note says that 2,400 were lifted and 2,350 were placed into baskets, its immediate accounting question is the disposition of the other 50. They may have been held aside or counted twice. The records must establish which explanation, if any, is supported.
Why surface area beats a casual tank label
Abalone attach to surfaces. A phrase such as 800 juveniles per tank gives almost no information about crowding unless tank geometry, shelter design and accessible grazing area are known. A two-metre tank with smooth walls is not equivalent to a tank furnished with double-sided shelter plates or perforated baskets. Likewise, the inner area of a basket is different from the footprint of the water vessel holding it. A report should state its density basis exactly: animals per square metre of accessible shelter, per square metre of internal basket surface, per litre of water, or per complete rearing unit. These denominators are not interchangeable.
Measure the usable surface according to the local protocol, subtracting regions that animals cannot reach or cannot safely occupy. In a teaching example, six baskets have 0.75 square metre of documented accessible area each. Their total area is 4.50 square metres. If 2,700 confirmed juveniles enter, the calculated initial density is 2,700 divided by 4.50, or 600 juveniles per square metre of accessible area. That is a description of the example, not a recommendation to stock at 600. If the area estimate excludes undersides that animals actually use, or includes surfaces blocked by frame members, the denominator must be corrected before comparing baskets.
Stocking calculation with unequal baskets
Imagine a nursery with three receiving baskets. The first has 0.60 square metre of usable surface and takes 330 juveniles; the second has 0.85 square metre and takes 425; the third has 0.75 square metre and takes 390. The local densities are 550, 500 and 520 juveniles per square metre respectively. The combined inventory is 1,145 animals and the combined area is 2.20 square metres, giving 520.45 animals per square metre. That overall figure is a ratio of totals, not the simple average of the three density figures. The latter would describe an unweighted average basket, a different statistical question.
This distinction matters when a facility changes its shelter layout. A report may show a lower animal-per-tank number after a transfer while crowding on the actual habitat surfaces has increased. It may also show the same overall density while one basket is dangerously overloaded and another is underused. A translator should preserve both the facility-level aggregate and the distribution among enclosures. When the intended comparison is experimental, report the replicate units, their individual outcomes and the planned analysis instead of treating the hundreds of animals in one shared tank as independent environmental replicates.
Detaching and transferring without inventing recovery
Moving juvenile abalone is not equivalent to pouring an exact number of beads between containers. Animals adhere, hide in grooves and may be injured during removal. A plate census taken before collection therefore has a different evidential status from a receiving-basket count after collection. Record the collection method, whether a count was direct or estimated from a sample, the time outside water, signs of shell damage and any animals kept for assessment. Translators should not replace a cautious source phrase such as recovered alive with transferred successfully if the destination count has not yet been confirmed.
Suppose staff estimate 3,000 juveniles across settlement plates but recover only 2,760 alive, including 60 segregated for observation. A statement that nursery intake was 2,760 is inaccurate if only 2,700 entered standard nursery baskets. The correct record distinguishes estimated available 3,000, recovered alive 2,760, standard nursery entrants 2,700 and separate observation holding 60. The difference of 240 between estimate and recovery is not an observed death count. It is an unverified gap until the source plates, discard containers and counting process are reconciled.
Transfer stress and the time window of survival
A seemingly normal count immediately after transport does not establish that juveniles tolerated handling. Some losses appear during the following days when stressed animals fail to attach securely or resume feeding. Choose the assessment window before interpreting survival. A same-day receiving count, 24-hour attachment check, seven-day short-term survival and 90-day nursery outcome are different metrics. Their numerators may refer to different categories of animals and their denominators may be initial transferred animals or animals that passed an early screen. Each definition must travel with its percentage.
For a hypothetical transfer of 2,700 juvenile abalone, 2,620 are confirmed alive and attached after seven days. Seven-day transfer survival is 2,620 divided by 2,700 times 100, or approximately 97.04 percent. If a later nursery count is 2,400, whole-period apparent survival measured against the original 2,700 is about 88.89 percent. The second calculation cannot be described as survival from day seven unless its denominator is changed explicitly to 2,620, which would yield approximately 91.60 percent for that later interval. Mixing the denominator changes the story.
Biofilm and the change of feed
Newly settled abalone may graze on films containing benthic microalgae, including diatoms, on their attachment surfaces. As the animals grow, nurseries may introduce larger seaweed pieces or formulated feed appropriate for the species and system. The transition is not completed by the first day that seaweed appears in a tank. The meaningful evidence includes bite marks or feeding observations, disappearance of appropriately offered food, growth, digestive condition and subsequent survival. Describing a feed as accepted requires a defined criterion; describing a tank as fully weaned may require proof that the new diet supports the entire graded group.
SEAFDEC/AQD describes diatom slurry feeding for early H. asinina juveniles and later nursery feeding with seaweed or formulated feed. That is a useful reference for stage sequencing, not authority for a universal feeding rate. A culture note should state the macroalgal species or diet formulation, whether its mass was measured wet or dry, any washing or draining procedure and how leftover material was counted. A wet mass of seaweed includes substantial water, so its apparent disappearance cannot be interpreted as dry-nutrient intake. Different harvest hydration or surface moisture can distort comparisons between feed lots.
Feed offered is not feed ingested
An elementary feed ledger begins with mass offered and mass recovered. The apparent mass disappearance equals offered minus recovered. If a basket receives 4.0 kilograms of wet seaweed and staff remove 1.3 kilograms of visible remnants, the recorded disappearance is 2.7 kilograms wet mass. It is not necessarily 2.7 kilograms eaten: portions can fragment, leach into the water, lodge in drains or change water content. A translator should retain the technical label apparent disappearance or apparent consumption when losses outside ingestion have not been independently quantified.
A more careful trial weighs a blank feeder exposed to the same water without animals and uses controlled draining times to estimate non-feeding loss. Even that method may not capture all differences between baskets. If the blank loses 0.4 kilograms from an initial 4.0 kilograms over the observation window, it signals a potential process loss; it does not automatically authorize subtracting exactly 0.4 kilograms from every live basket. Control variability, water flow, algal form and animal activity must be considered. When the document has no controls, avoid reporting precise intake efficiency or feed-conversion claims from disappearance alone.
Wet and dry basis are different quantities
Wet seaweed and formulated feed cannot be compared meaningfully without defining moisture basis. Suppose a wet feed sample is 12 percent dry matter. Four kilograms of that material contain 0.48 kilograms of dry solids and 3.52 kilograms of water by mass. If a formulated feed is 90 percent dry matter, four kilograms of it contain 3.6 kilograms of dry solids. The two four-kilogram offers differ dramatically in actual solids. A sentence that one feed required less mass might simply be reporting moisture content rather than a biological improvement.
To convert, multiply wet mass by the dry-matter fraction obtained from a representative sample measured by an appropriate method. To convert a dry-matter amount back to expected wet mass, divide by the fraction, not subtract it. Be careful when a feed table uses moisture percent as a percentage of wet weight while another uses water-to-dry-solids ratio. These are different mathematical definitions. If the material is salt-laden or retains surface seawater, the measured solids may include inorganic salt as well as nutrients; dry matter is not identical to digestible organic matter.
Shell length does not tell the full growth story
The easiest visible size measurement is usually shell length, but length is not body mass and a high average shell length does not guarantee a narrow cohort. Define whether length means maximum shell dimension, the instrument used, the position of the caliper and how damaged shell margins are handled. A nursery report may include total animal weight, drained wet body weight, shell length or an inferred weight from a size relationship. Translating one variable as another changes the scientific claim. Species, water retention and measurement procedure should accompany every observed weight.
For a simple example, a cohort’s sampled average shell length moves from 13.0 millimetres to 24.0 millimetres over 77 elapsed days. The difference is 11.0 millimetres, and the average linear increase is 11.0 divided by 77, approximately 0.1429 millimetres per day. This is not a universal individual growth rate; it is a comparison between two sampled cohort summaries. If the initial and final samples contain different animals, size-selective deaths or transfers may change the mean even without equivalent growth in every survivor. Preserve that caveat.
Use the right temporal denominator
Growth per day can be calculated using calendar days, days actively reared in the nursery or a thermal-time method, depending on the question. The denominator must be explicit. A report that counts eight weekly sampling visits and divides the length gain by eight is reporting gain per interval, not gain per day. A facility may also pause growth measurements during quarantine. Such a pause does not remove the elapsed time from a biological growth-rate denominator unless the study defines and justifies a different clock. State actual dates rather than writing vague phrases such as after two months.
If a group grows from 14 to 26 millimetres between 2 June and 23 August, calculate the elapsed days from the recorded timestamps rather than multiplying an assumed four weeks by three. A difference between inclusive day counting and elapsed-time counting can alter the denominator by one day. For short trials, that matters. A translator faced with only month names should avoid introducing an exact daily rate. It is better to say that shell length increased by 12 millimetres over the stated observation period and request dates than to manufacture precision.
Sample a cohort rather than an attractive corner
Juvenile abalone do not necessarily distribute themselves uniformly across a basket. A sampler may favor the exposed surfaces, larger animals and accessible corners. If the question is the population’s size distribution, define a selection plan that gives the hidden and visible animals a fair chance to be counted. Randomize plate order, sampling locations or basket selection as appropriate, and document deviations. The number of animals measured is not itself a guarantee of representativeness. A very large convenience sample may be less reliable than a smaller, well-designed probability sample.
When reporting a mean shell length, include sample size, a measure of spread and the sampling frame. For example, a mean of 22.4 millimetres from 80 selected juveniles without a standard deviation cannot establish that nearly all animals exceed 20 millimetres. If those 80 came from the most visible plate, the report does not represent the entire tank. A useful translation preserves phrases such as sampled on upper trays or measured from three randomly chosen baskets. Those few words often control how broadly a numerical result may be interpreted.
Weighted mean versus mean of groups
Suppose Basket A has 20 measured juveniles with mean length 24 millimetres, while Basket B has 80 with mean length 20 millimetres. The pooled sample mean is (20 times 24 plus 80 times 20) divided by 100, or 20.8 millimetres. The simple average of basket means is 22 millimetres, but it treats a twenty-animal basket sample as equally influential as an eighty-animal sample. Neither number is automatically a true whole-facility mean if the samples were drawn unequally relative to basket populations.
The right estimator depends on the sampling design. If each basket is intended as an equally weighted experimental replicate, a mean across baskets may be appropriate for comparing enclosure effects. If the target is the average length of a randomly selected animal from all baskets, design weights reflecting eligible animal numbers are needed. Report what was averaged and why. A translator should never turn an average of basket means into an average individual length without checking the plan. The correct label is often more important than adding another decimal place.
Size grade distributions and transfer gates
A grow-out facility may accept advanced juveniles only above a local shell-length threshold, while a nursery keeps smaller animals for additional rearing. A mean size is insufficient to establish readiness. Imagine 1,000 measured animals distributed as 180 below 20 millimetres, 470 between 20 and below 25, and 350 at or above 25. A statement that the cohort is 25-millimetre seed would be misleading even if a few animals exceed that mark. The proportion at or above 25 is 350 divided by 1,000, or 35 percent of the measured sample, not necessarily 35 percent of the unmeasured inventory.
Grade boundaries should specify inclusive or exclusive endpoints and the measurement protocol. If a buyer uses greater than or equal to 25 millimetres, an animal recorded as exactly 25.0 counts in that band. If another specification says greater than 25, it does not. Rounding can move borderline cases across a threshold, especially when calipers are recorded only to the nearest millimetre. Preserve the source precision and do not infer exact threshold compliance from a rounded average. The appropriate next action may be regrading, stratified holding or a documented buyer acceptance test.
Regrading changes the composition of cohorts
A nursery may separate faster-growing juveniles from slower-growing ones. After grading, Tank A can show a higher mean size, not because its residents suddenly grew, but because it received large individuals from several baskets. The old cohort has been partitioned into branches. To translate a performance summary, identify whether it describes biological growth within one traced group or an administrative selection of animals. A post-grading size mean cannot be compared directly with a pre-grading mean as an individual growth estimate without accounting for entry and exit.
For example, an original group of 2,400 animals is divided into 900 advanced juveniles, 1,350 retained smaller animals and 150 requiring separate observation. The distribution balances exactly, but only if every category uses the same time point and none is a duplicate count. If 20 animals were discarded because their shells were badly damaged, they must appear as a separately documented disposition, not hidden inside the retained group. Give each branch a child identifier and maintain the links to the original parent cohort. Later nursery survival must not count the same animal in two child branches.
Mortality and missing animals
Dead shells, missing animals and live animals placed in quarantine represent separate accounting states. A cleaning record may list 45 shells removed over a month, but that is not necessarily the complete mortality for the period. Soft tissues may disappear, empty shells may remain from an earlier interval, and small live juveniles may hide under fittings. Report confirmed mortality separately from probable loss and unresolved count difference. Translating an inventory shortfall as dead animals introduces a causal claim that the evidence does not support.
Suppose opening stock is 2,700, confirmed incoming transfers are 150, confirmed outgoing transfers are 200 and 80 deaths were recorded. Expected closing stock equals 2,700 plus 150 minus 200 minus 80, or 2,570. A physical sample-based estimate suggests 2,530 present. The 40-animal difference is an unresolved discrepancy, not an additional 40 confirmed deaths. A conservative report could say that confirmed mortality was 80 and closing inventory was 40 below the ledger expectation, with method and uncertainty stated. Only further evidence might connect that discrepancy to mortality, escape or counting bias.
Survival by interval and cumulative survival
If a batch starts with 2,700 juveniles, has 2,430 verified survivors after the first nursery interval and 2,187 after the second, first-interval survival is 2,430 divided by 2,700, or 90 percent. Second-interval survival is 2,187 divided by 2,430, also 90 percent. Cumulative survival from the beginning is 2,187 divided by 2,700, or 81 percent. Adding the two interval survival percentages or claiming a 90-percent total survival would be wrong. Survival probabilities multiply when the periods are nested and the cohort accounting is sound.
The calculation becomes more complicated when animals are added or sold during the interval. A simple end-over-start ratio can then be misleading. One practical remedy is to report survival by uniquely traced cohort and interval, with entry and exit events logged separately. If the facility uses animal-days or a time-at-risk model, explain its assumptions instead of equating the result to a simple headcount ratio. A sentence such as 92 percent remained should specify remained from what starting population, at which checkpoint and whether graded or transferred animals were deliberately removed from observation.
Water quality as context, not a single magic number
Temperature, salinity, dissolved oxygen, water movement and accumulated wastes influence juvenile performance, but one isolated reading does not establish that conditions were safe throughout the period. A tank’s morning oxygen observation can miss a low point overnight. A measured salinity may represent the inlet rather than a poorly flushed basket interior. Translate water-quality records with units, sampling location, timestamp, depth and instrument method. Where a sensor was not calibrated, the value should be qualified accordingly rather than reported as an unquestionable property of the entire facility.
Species-specific guidelines are essential. A source describing one tropical species under one feed and flow-through system does not make those numeric settings generally safe for every abalone. Compare actual measurements against the facility’s validated operating range for the identified species, life stage and equipment. If a report gives dissolved oxygen as 6.1 mg/L, retain both the unit and the time of the measurement. Do not translate it as 6.1 percent saturation, a completely different metric whose numerical scale depends on temperature, salinity and atmospheric pressure.
Flow-through water and recirculation
Two nurseries may use identical tank volumes yet deliver very different water renewal and waste removal. In a flow-through system, the important descriptors include inlet flow, tank mixing, bypass, actual residence time and the condition of the influent water. In a recirculating system, total circulation flow is not equal to net new-water addition. The distinction matters because a brochure may call both systems continuously supplied with clean water. Translate the mechanism rather than treating the word continuous as proof of effective exchange or waste removal.
For a nominal 1,500-litre nursery tank receiving 75 litres per minute of new water under steady operation, nominal hydraulic residence time is 1,500 divided by 75, or 20 minutes. That is not proof that every litre is replaced in 20 minutes or that there are no stagnant zones. If only 10 litres per minute is new makeup water and 65 litres per minute is recirculated, dividing volume by 75 describes circulation turnover, not new-water residence. In a reliable record the two flow streams have separate labels, and any uncertainty in flowmeter calibration is documented.
Feed transition diary and observation criteria
A strong feeding diary records what feed was offered, how it was prepared, when it was presented, when leftovers were removed, and how the juveniles behaved. A note saying switched to seaweed on Monday may mean trial introduction, partial replacement or complete withdrawal of the biofilm diet. Translators should identify the operational criterion used to call the transition complete. If a facility reports only the day of the first seaweed offer, label the event accurately and do not infer that all animals were feeding on it by the following day.
A practical monitoring sheet can include observations of active grazing, attachment stability, visible uneaten food, fouling, stress, size variance and the number of animals sampled. Do not use a single clean-looking tray as a proxy for all enclosures. If 60 of 80 sampled juveniles show evidence of the designated feed in a defined assessment, the observed sample percentage is 75 percent. It is not automatically the share of the entire nursery that has been successfully weaned. The evidence may support a cautious statement that feed acceptance was observed in the measured sample under the stated criterion.
Recognize how feed and water alter each other
A large wet seaweed offer can serve as food and also as a physical object that changes flow paths, shelter access and waste accumulation. Feed pieces may decay before the next cleaning, while formulated particles can collect underneath inaccessible surfaces. A feeding comparison should therefore record maintenance practice and water regime alongside diet type. If a growth difference appears after both feed and flow were changed, the report cannot isolate the causal effect of diet. Distinguish operational association from a controlled feeding experiment in translation.
Similarly, a statement that no food remained at the end of a week can mean that animals consumed the ration, it washed through the mesh or staff removed it during unrecorded cleaning. Production records should use consistent collection times and disposal notes. For a formal feed-conversion analysis, separate the estimated mass ingested from mass offered, and specify wet or dry feed basis. The reader needs to know whether numerator losses and denominator growth refer to the same animals and period. Otherwise a calculated conversion ratio may look exact while comparing quantities that were never measured together.
Quality control of length and count
Measurement errors occur when calipers are misread, animals are measured at different body orientations or the form rounds all lengths to the nearest five millimetres. Build a small duplicate-measurement exercise into the nursery record. If two observers differ systematically, agree the anatomical landmarks and repeat training before interpreting shifts of a fraction of a millimetre as biological progress. The aim is not to force every pair of readings to match; it is to understand the measurement repeatability so that a reported growth difference is larger than the noise when strong claims are made.
Counting errors can have different directions. A plate with concealed juveniles may be undercounted; an animal seen twice while moving between adjacent surfaces can be overcounted. A subsample estimator needs a defined total sampling frame, selected units and variance estimate. If the method changed from whole-container enumeration to photographic counts mid-trial, describe that change. A trend in the count could partly reflect changed visibility or detection. Never use a single high-precision calculator output to conceal a method that produced only approximate observations.
A complete worked nursery ledger
Consider a training cohort of 3,000 juveniles estimated on source plates. Of these, 2,800 are directly recovered and verified alive, 50 are assigned to a separate observation container, and 2,750 are admitted to standard nursery baskets. Over the first 30 days, the nursery documents 110 confirmed deaths and transfers 200 living juveniles to a grading study. The expected day-30 closing stock is 2,750 minus 110 minus 200, or 2,440. A count of 2,420 after an access-limited inspection leaves 20 unresolved, subject to the uncertainty of the inspection. It would be incorrect to claim precisely 330 deaths since the source-plate estimate.
Using standard nursery entrants as denominator, the observed confirmed death fraction by day 30 is 110 divided by 2,750, or 4 percent. The remaining percentage cannot simply be called 96 percent alive in the original nursery because 200 animals were intentionally transferred out and 20 are unresolved. A better dashboard displays confirmed deaths 110, outbound living 200, expected retained 2,440, observed estimated retained 2,420 and unresolved difference 20, with dates. The separate observation container is another branch with its own count history. Clear categories are more valuable than one appealing percentage that hides movements.
An end-of-nursery size example
Assume the original eligible nursery stock was 2,750 and after 80 days a verified final inventory across all traced standard branches is 2,310 living animals, excluding any explicitly transferred research group that remained alive elsewhere. Sample the final population with a defined design and obtain an average shell length of 26.2 millimetres, a sample standard deviation of 3.1 millimetres, and 240 measured animals. The values describe the measured group and its dispersion. The report should not say that all 2,310 are at least 25 millimetres. The sample would require its actual grade counts to support that claim.
Suppose 164 of the 240 sampled animals meet a stated threshold of at least 25.0 millimetres. The observed proportion is 68.33 percent, but uncertainty and sampling design matter before projecting it to the full inventory. If the sample is representative, multiplying 2,310 by 164 divided by 240 yields an estimate of 1,578.5 qualifying animals, not a direct count of 1,579. Report it as an estimated qualifying number with appropriate uncertainty, not an exact number ready to ship. A final shipping count must come from actual grading and dispatch records, including any damaged or quarantined animals.
Independent transfer problem: density
Try this without looking at the answer. Four nursery baskets have usable areas of 0.55, 0.65, 0.70 and 0.60 square metres. They receive 275, 390, 420 and 330 confirmed juveniles respectively. Calculate the density in each basket, the facility total density, and the risk of summarizing the farm as if every basket held the same crowding. Then ask what other information you would need before deciding whether these densities are acceptable. A correct arithmetic result is only the beginning; biological acceptability depends on the specified species, stage, available surface, water and feeding management.
Answer: the four local densities are 500, 600, 600 and 550 juveniles per square metre. Total animals are 1,415 and total usable area is 2.50 square metres, so combined density is 566 per square metre. The simple average of local densities is 562.5 per square metre, not the ratio of overall totals. More important, a combined figure conceals the two baskets at 600. Before judging suitability, verify how area was measured, the size distribution, feed availability, flow and applicable protocol. No universal safe stocking density follows from this calculation alone.
Independent transfer problem: survival
A nursery has 1,800 confirmed entrants. At day 20, it documents 90 deaths and 110 live transfers to another tracked unit. Its physical census estimates 1,590 animals remaining. Calculate expected retained stock, discrepancy and confirmed mortality fraction relative to the original entrants. Then decide whether the recorded evidence supports calling the discrepancy extra deaths. Finally explain how the ratio changes if 40 animals were later discovered alive in the filter housing and assigned back to the same cohort. Be explicit about dates because a later finding must not be silently moved backward in time.
Answer: expected retained stock is 1,800 minus 90 minus 110, or 1,600. The day-20 estimate of 1,590 is short by 10 animals, and confirmed mortality is 90 divided by 1,800, or 5 percent. The ten missing animals are not confirmed dead. If 40 live animals are later found, the census method or movement history needs investigation rather than simply claiming a negative mortality rate. Record when and where they were discovered, check whether they were previously included in another branch and reconcile the parent ledger. This is precisely why an observed inventory and a ledger expectation are distinct fields.
Independent transfer problem: size and feed
A representative size sample rises from a mean of 15.5 millimetres to 25.7 millimetres over 85 elapsed days. Calculate the difference and apparent mean linear growth per day. Separately, a nursery offers 18 kilograms of wet seaweed over a trial and collects 5 kilograms of wet remnants. Calculate apparent disappearance, then list three reasons why that quantity is not actual ingestion. If a report calls both results individual growth and feed efficiency, explain what information is missing. Try to answer before reading the next paragraph, because the missing assumptions matter as much as the arithmetic.
Answer: mean length difference is 10.2 millimetres and the apparent mean rate is 10.2 divided by 85, or 0.12 millimetres per day. Apparent wet seaweed disappearance is 13 kilograms. It may include fragmentation, leaching or waste washed out of the system, and the moisture content of offered and recovered material may differ. The mean length change may also be affected by selective survival or unequal sampling. Actual individual growth requires linked repeated observations or a defensible population design; feed efficiency additionally requires verified intake basis and an appropriate comparable biomass gain denominator.
Uncertainty is a property of the evidence
Do not turn a source estimate of roughly two thousand juveniles into an exact English number of 2,000 simply because it fits a spreadsheet cell. Distinguish directly counted, estimated by subsampling, inferred from mass and reconstructed by ledger balance. For subsampling, the uncertainty depends on heterogeneity across units, the selection design, observer reliability and the number of independent units. Repeating a hundred counts within one small patch does not substitute for sampling across the whole nursery. A translation can preserve uncertainty without becoming cumbersome by adding concise labels such as estimated, based on 12 plates, or subject to recount.
A percentage computed from uncertain numerator and denominator should not have more meaningful precision than those inputs warrant. If approximate counts are 2,750 and 2,310, reporting survival as 84.0000 percent suggests unjustified precision. It may be reasonable to communicate approximately 84 percent provided the cohort boundaries and dispositions are appropriate. For formal analysis, retain raw counts, sampling errors, exclusions and the full calculation separately. An editor who improves readability by erasing the word approximate may change the scientific conclusion, especially when the value is close to a release or contractual threshold.
Translation errors and repair workflow
Common errors have recognizable patterns: translating nursery intake as hatchery production, calling a plate density a water-volume density, replacing wet seaweed mass with dry feed mass, turning feed disappearance into consumption, and confusing mean shell length with the number above a grade boundary. Another frequent error is a hidden unit shift from millimetres to centimetres. A 25-millimetre juvenile is 2.5 centimetres, not 25 centimetres. Translated tables should preserve units in headings and verify that any converted values and thresholds are internally consistent.
Repair these problems in a fixed order. First copy the observed term and its source-language unit. Second identify the measurement event and the physical denominator. Third calculate an independent check on a small example. Fourth compare the result with source definitions and replicate records. Fifth replace any overconfident causal claim with language supported by the measurement. Finally read the entire report for consistency: stock numbers should reconcile, size grades should sum to the counted total, and a transfer should appear as outbound in one unit and inbound in another. Do not solve an inconsistent ledger by inventing a missing death count.
Reporting across languages and audiences
Technicians, buyers and public readers need different amounts of detail, but they need the same underlying meanings. A technical table may retain cohort identifiers, metric definitions, sampling design and uncertainties; the accompanying plain-English explanation can say that some juveniles were moved to another unit rather than lost. For international readers, expand abbreviations on first mention and distinguish stocking density, survival percentage, growth rate and feed ratio. Do not use seed as an undefined synonym for egg, embryo, settled larva and juvenile. In abalone aquaculture the word seed usually refers to young animals intended for further culture, but its size and readiness remain protocol-specific.
Maintain a translation glossary with stage name, original term, preferred English label, unit, measurement method and an example. Include terms such as attached, recovered alive, accepted into nursery, sampled, retained, graded out and released for grow-out. A bilingual colleague should be able to follow an animal through these events without guessing whether a percentage uses estimated plate stock or verified nursery intake. Whenever the source lacks a field, use not reported rather than filling it from another species’ manual. Clear omissions are often safer and more useful than apparently complete but invented data.
FAQ: when is an abalone nursery batch successful?
Success depends on a defined operational objective, not a single survival percentage. An intermediate nursery may be tasked with turning verified early juveniles into a specified size grade by a target date while maintaining animal health and traceability. Report the number actually meeting that grade, the number still alive but below it, animals under observation, confirmed mortality, estimated losses and production time. If the batch achieves a high apparent survival but almost none are ready for the next stage, the result is not equivalent to a high ready-to-ship yield. The metric should follow the management question.
A report may also discuss animal welfare, disease screening, food source, energy use and environmental performance. Those are related but distinct dimensions. Do not translate a production yield into a blanket claim that a facility is sustainable or disease-free. Confirm the relevant evidence and standards before using stronger labels. When comparing two systems, check species, starting size, stocking surface, diet, water regime, duration, grade threshold and accounting conventions. Without those controls, one facility’s 90 percent survival and another’s 85 percent could reflect different denominators rather than a genuine performance difference.
FAQ: can we infer survival from mean length?
No. An increase in mean shell length is compatible with good individual growth, but it can also occur when smaller animals die or are moved. A nursery could lose many undersized juveniles and still show a rising average length among the remainder. Survival requires count-based or otherwise defensibly estimated living-animal data tied to a cohort and time interval. Size information is valuable for grading, but it cannot replace the movement and mortality ledger. A full report should show both numbers with their methods and, where possible, distributions rather than only means.
Nor does a stable mean prove growth has stopped. A group may have some animals growing quickly while smaller entrants arrive, keeping the pooled mean approximately unchanged. Grading events can distort the picture in either direction. To investigate, trace individual or well-defined subgroup histories, measure a representative distribution, and compare the same grade boundaries over time. The correct English explanation might say mean size of animals present was stable, rather than the juveniles did not grow. That small change preserves the distinction between a measured summary and a claim about each animal.
FAQ: does a good-looking feed ratio prove efficiency?
Not without a clearly defined ratio. A report might divide fresh wet seaweed offered by live-weight gain and call it feed conversion; another may use dry formulated feed consumed; a third may count only purchased feed, excluding harvested on-site algae. These ratios have different meanings and cannot be compared directly. Confirm the numerator’s feed basis, the denominator’s animal gain basis, period dates, mortalities and removals. Even when each ledger is consistent, conversion can be affected by temperature, initial size, species and leftover feed loss. A bare number should never be detached from its definition.
In an ordinary nursery, tracking feed use can still be operationally helpful without implying a controlled efficiency experiment. A manager may use offered feed per thousand surviving juveniles to plan supply, while a scientist needs intake and biomass gain for a physiological question. Translate the purpose explicitly. Do not write that a hatchery discovered a better diet if the data only show lower purchased-feed expenditure following a move to locally collected seaweed. That is an economic observation subject to labor, harvesting costs and nutritional context, not automatically improved feed conversion.
A compact field checklist
Before a nursery record is considered ready for public explanation, verify the following information in prose or a well-labeled table: species, cohort origin, receiving date, verified entrants, destination containers, accessible surface and method of measuring it, interval survival definition, mortality and transfers, feed identity and mass basis, sampling procedure, shell-length distribution, final grade counts and unresolved discrepancies. Add water-quality context and a note on deviations from the normal operating procedure. If one critical field is absent, say what cannot be concluded rather than inferring it from the outcome.
A good final summary is deliberately narrow. For instance: Of 2,750 confirmed juveniles admitted to the nursery, 2,310 were verified alive across traced standard branches after the stated period; a representative sample averaged 26.2 millimetres shell length, and the number meeting the grow-out grade remains an estimate pending full grading. That sentence is less dramatic than saying 2,310 market-ready abalone were produced, but it is useful because every claim can be followed back to a record. The purpose of technical translation is to protect that traceable meaning.
Sources and verified neighboring guides
SEAFDEC/AQD: Farming Abalone — nursery and grow-out guidance
FAO: Biology and Culture of Selected Species, abalone stage and density context
Philippine Journal of Science: diatom diets and attachment substrate for Haliotis asinina
NOAA Fisheries: abalone juvenile grow-out assembly and conservation context
eduKateSG: preceding abalone settlement and metamorphosis guide
eduKateSG: preceding abalone broodstock and fertilisation guide
eduKateSG: aquaculture RAS turnover and residence time guide