Spiny lobster puerulus nursery records translate into useful decisions only when each batch has an intact chain of counts. Start with the number of living seed lobsters accepted, record losses during holding and transport, identify exactly when animals become juveniles, and calculate survival through each interval. A batch that arrives alive can still suffer delayed losses, so arrival survival and nursery survival should never be treated as the same outcome.
To understand a report, ask which species is involved, whether seed came from a lawful wild collection or a documented hatchery, and how many animals entered each defined period. The survival proportion for an interval is the living animals at the end divided by the eligible animals at its start, with any additions, removals and transfers accounted for. If animals are divided among cages, reconstruct a batch ledger instead of adding convenient percentages.
The quick method is: identify species and stage; record legal seed provenance; audit received and transferred counts; monitor holding time and water handling; separate immediate from delayed mortality; record moulting and transition to juveniles; use stage-specific start and end denominators; distinguish disappearance from confirmed death; calculate cumulative survival; and connect any husbandry claim to actual matched evidence.
Where the nursery phase begins
A spiny lobster’s phyllosoma stage belongs to its offshore larval development. The puerulus is the transitional post-larval stage before the pigmented benthic juvenile. Many farming systems obtain pueruli from licensed wild-seed fisheries rather than a fully commercial closed-cycle hatchery, although technical progress in hatchery production has been made. The source should be recorded honestly. A purchase record saying hatchery seed is not proof that a facility reared larvae through the long phyllosoma sequence; the supplier and provenance require evidence.
The puerulus transition is a vulnerable operational period because the animal is small, handling may be frequent and feeding behaviour changes as it becomes a juvenile. Scientific sources describe the tropical industry’s nursery as a separate bottleneck, with substantial variation in survival. Do not generalise a historical field percentage as a universal rate for all species, sites and facilities. Improved practice is evaluated through comparable cohorts, not the repetition of an attractive number from a different country or decade.
A nursery may receive animals directly from collectors, through a dealer or from a controlled breeding facility. Each step creates an opportunity for unrecorded time, handling and animal loss. A complete batch identity carries through collector, dealer, transport, arrival tank, acclimation, early holding, first juvenile moult and transfer to grow-out. The dates need consistent local time and the counts must specify which events are physical counts and which are estimates.
Keep the life-stage terminology in the original report. If a nursery receives 800 pueruli and later records 620 first-stage juveniles, the number 620 is not automatically a survival estimate unless all movements and counting conditions are known. Some animals may still be pueruli, some may be hidden in shelters and some may have been transferred. A label that says juveniles recovered is a count of recovered juveniles, not necessarily a census of everyone still alive.
A ledger that can be balanced
For each batch, record the number of living animals received, dead on arrival, uncertain-condition animals, animals transferred in or out, animals found dead later, unaccounted animals and those counted alive at each checkpoint. The fundamental inventory relation is opening animals plus additions minus known removals equals closing animals plus confirmed deaths plus other reconciled changes. When numbers do not balance, flag a discrepancy instead of inventing an explanation.
A batch of one thousand animals is checked on arrival: 930 are alive, 45 dead and 25 unaccounted because a packing container leaked. The first reported survival among all thousand shipped is 93 per cent, but the status of 25 must be investigated. Treating the 25 as confirmed deaths changes the mortality classification even if the final live count is unchanged. Treating them as having survived would be worse. Good evidence distinguishes observed loss, inventory discrepancy and known physical cause.
A subsequent seven-day count finds 870 animals alive, with sixty confirmed deaths and no additions, removals or escape. Conditional seven-day survival is 870 divided by 930, approximately 93.55 per cent. Overall survival from the shipped thousand to seven days is 870 divided by 1000, or 87 per cent. The conditional and cumulative denominators differ. The report should display both instead of saying the nursery had 93 per cent survival without defining when the clock started.
When cohorts are split into cages, use counts for pooled survival. If cage A starts with 100 animals and ends with 90, its survival is 90 per cent. Cage B starts with 900 and ends with 630, giving 70 per cent. The simple average of 90 and 70 is 80 per cent, but total surviving animals are 720 of 1000, or 72 per cent. Whether a cage-balanced summary or animal-weighted summary is relevant depends on the management question. The author needs to say what was averaged.
Transport is a distinct exposure period
Transport begins before the first vehicle moves. Collection, temporary holding, sorting, bagging, oxygenation, loading and unloading can all change the animal’s condition. A transport record therefore requires the first time of confinement, origin, batch code, species, packing method, holding volume or vessel details, temperature monitoring, destination and time of opening. Distinguish observed measurements from a handler’s recollection when records are incomplete.
Historical research and field reports from Indonesia describe improvements from dry sand-based transport towards water-based methods with aeration or oxygen provision for seed lobsters. That experience is context, not a universal do-it-yourself packing formula. What matters for interpretation is whether the actual method maintained conditions appropriate to the species and stage for the specified travel duration. Follow qualified husbandry guidance and local animal-welfare requirements rather than applying arbitrary numbers from a case study.
The words aeration and oxygen supply are not interchangeable. Aeration exchanges gas with surrounding air and may also aid mixing, depending on the vessel design. Compressed oxygen can change available gas supply but does not eliminate all water-quality problems. Carbon dioxide accumulation, temperature changes, ammonia, crowding and rough handling can still contribute to stress. A log that records only oxygen added cannot prove that conditions remained suitable throughout a long journey.
Temperature logs need timestamps, sensor location and calibration. A single maximum reading might come from a briefly exposed external probe rather than the water actually surrounding animals, or it might identify a true dangerous excursion. Compare sensor placement and sampling interval, reconstruct duration above any scientifically justified operational alert level, and examine subsequent outcomes. Do not treat a threshold borrowed from an unrelated species as an established physiological limit for P. ornatus or P. homarus.
Arrival counts are not the finish line
On arrival, a technician may see animals swimming and mark all of them alive. Later, some may become inactive or die during acclimation. If a report asks whether a transport method was successful, the observation window must be specified. Arrival survival, twenty-four-hour survival and seven-day survival measure different things and can change the apparent ranking between two methods.
Describe compromised condition separately from confirmed mortality. A weak animal may recover or die; its state should be documented under an agreed observational rubric. Observers should not casually use clinical disease terms based only on visible stress. Disease confirmation may require laboratory diagnosis, and healthy-looking animals can still harbour pathogens. A survival ledger can identify a problem without proving its pathology.
A reliable comparison of two transport approaches requires balanced batches of similar stage and species, comparable starting condition, temperature and travel time, and adequate replication. If method A was used only on a short cool journey while method B was used only on a long hot journey, a simple survival contrast is confounded. The honest reading is that performance differed under different conditions; the causal effect of packing method remains unresolved.
Metamorphosis and first juvenile counts
Puerulus-to-juvenile transition changes the question from survival alone to developmental progress. A record may report living pueruli, juveniles after the next moult, animals in the process of moulting and unknown animals hidden in shelters. A transformation percentage needs a denominator: juveniles observed divided by the original living pueruli at a specified start, or juveniles divided by animals known to survive to a particular date. These give different quantities.
A batch with 600 living pueruli at a checkpoint, of which 420 have become juveniles, 120 remain pueruli and sixty are confirmed dead, contains 420 juvenile transitions as a proportion of 600 starting animals, or 70 per cent. Survival regardless of stage is 540 over 600, or 90 per cent. Among the 540 living animals, 420 divided by 540, or 77.8 per cent, are juveniles. Calling all these values survival is inaccurate. State transition, conditional stage composition and survival separately.
Moult records can be misleading because discarded shells are not dead animals. An exuvia is evidence that a moult happened, but it is not always straightforward to assign the shell to a particular individual. In communal tanks, a raw shell count should not become a count of distinct juveniles without validation. Inspectors may need careful animal counts, a consistent classification protocol and appropriate handling to avoid mistaking moults for mortality.
Shelter, density and the limits of causal language
Juvenile spiny lobsters use shelter, and nursery studies have examined shelter materials, feeding levels and survival. A recent study catalogued in FAO AGRIS compared bundled-net and pipe shelters, as well as feeding rates, in P. ornatus experiments. Such experiments provide useful methodological evidence, but their treatment-specific results should not be presented as universal recipes. Animal sizes, tank designs, numbers of replicates, diets and statistical uncertainty need to travel with the conclusion.
Stocking density may be expressed per square metre of bottom area, per cubic metre of water or per physical enclosure. These are not interchangeable. Two tanks with the same water volume may have different floor areas and shelter arrangements. For a benthic juvenile animal, contact with shelter surfaces may matter more than a simple cubic volume suggests. Use the denominator actually stated in the original experiment and preserve it when comparing nursery records.
Crowding and loss can coincide without one observation alone proving cannibalism. Missing animals may have escaped, been concealed, been removed in cleaning or experienced predation. To attribute mortality to cannibalism, look for corroborating observations, experimental design and ruling out alternative inventory errors. Otherwise write that the losses were consistent with a hypothesis requiring investigation. The distinction protects both biological accuracy and farm decision-making.
Nutrition and water quality during transition
Nursery animals need stage-appropriate nutrition, but the earliest transition from puerulus to juvenile differs from later juvenile feeding. Some early stage observations involve a non-feeding puerulus phase, followed by increasing relevance of external feed as juvenile behaviour develops. Recording feed delivered is not the same as recording feed consumed. Uneaten fresh feed can deteriorate water quality, while pellets may behave differently. Any comparison should specify ration form, dry-matter or wet-mass basis and uneaten material handling.
Dissolved oxygen, temperature, salinity and other water variables should be logged with units and measurement context. A value without whether the sensor was calibrated, the water was mixed or the reading occurred before or after feeding can be hard to interpret. If losses cluster around a water-quality incident, establish chronology before making a causal claim. Keep alarms, corrective actions and animal outcomes in separate fields so the reader can see which events preceded which observations.
A successful nursery should be judged across several outcomes: survivors, developmental stage, size variation, condition, husbandry inputs and the ability to transfer animals safely to a later phase. A high apparent survival number based on animals seen near the front of a shelter does not prove that all animals were counted. Where individual counts are difficult, sampling methods and extrapolation assumptions need explicit documentation.
Uncertainty and honest comparisons
A reported survival proportion from one cage is an estimate for that observed cohort. It is not automatically the expected survival of every subsequent batch. Different starting conditions, season, operator skill, animal source, density and disease status can alter outcomes. Replicate units and repeated batches are needed to distinguish a robust improvement from ordinary variation. If the study has only one treatment tank and one control tank, unit-level confounding is a serious limitation.
Approximate binomial intervals can illustrate counting uncertainty when individuals are independently followed, but communal husbandry often violates this idealised assumption. If 90 of 100 individually independent animals survive, a point estimate is 90 per cent; the approximate standard error is square root of 0.9 times 0.1 divided by 100, or 3 percentage points. That calculation says nothing about shared water-system failure, correlated infections or miscounted hidden animals.
In practical writing, it is often better to state n of N survivors at a defined checkpoint and discuss the design than to add a falsely precise percentage. An intervention that improves survival in one specific nursery may be promising without being proven for all facilities. Keep the research question and uncertainty together rather than replacing them with a marketing statement about guaranteed yield.
Frequently asked questions
Is arrival survival the same as nursery survival?
No. Arrival survival measures the live animals at delivery relative to the shipped cohort. Nursery survival describes the living animals remaining after a defined subsequent rearing period. Delayed mortality means the two can differ significantly. Give each result its own start and end times.
Do dead-on-arrival and missing animals mean the same thing?
No. Dead-on-arrival indicates observed death, while missing means an inventory discrepancy or unobserved fate. Both may reduce live received numbers, but they imply different evidence about cause. Record them separately and reconcile quantities when possible.
Can percentages from successive stages simply be added?
No. Conditional stage survival proportions multiply to yield cumulative survival if the stages and denominators are aligned and all movements reconciled. A 90 per cent transport survival followed by 80 per cent nursery survival gives 0.9 times 0.8, or 72 per cent overall, not 170 per cent.
Does a high juvenile count prove a successful hatchery?
Not necessarily. The number might include wild-collected seed grown through a nursery phase. To claim hatchery production, there should be documented parental origin, egg or larval history and successful completion of earlier stages in a controlled facility. Nursery production and closed-cycle hatchery production are different claims.
Should an exuvia be recorded as a death?
No. An exuvia is a shed exoskeleton from moulting, not the body of a dead animal. Count mortality through appropriate inspection and retain separate moult observations. Identifying exuviae by individual may be difficult in a communal environment.
What is the safest way to interpret a transport trial?
Compare like animals under appropriately controlled and replicated conditions, record travel duration and environmental exposure, and evaluate a common survival window including delayed outcomes. If the journeys differed, describe the observed contrast without attributing all of it to the method.
Full worked ledger exercises: from collection to juvenile transfer
Exercise 1: survival on delivery
A fictional supplier seals 1,000 eligible pueruli for a journey. At receiving, the worker identifies 930 living animals, 45 confirmed dead and 25 animals not accounted for after one damaged container is inspected. Arrival survival relative to shipped animals is 930 divided by 1,000, or 93 per cent. Confirmed dead on arrival represent 4.5 per cent, while the unresolved discrepancy is 2.5 per cent. Those categories sum to 100 per cent but do not prove all missing animals died. The receiving report should list live, dead and missing separately, note the damaged container, and preserve the dispatch count and timestamp. A single phrase 7 per cent mortality would wrongly convert the unresolved group into observed deaths. The practical action is reconciliation, not a more elaborate formula.
Exercise 2: delayed loss after apparently successful transport
The 930 live arrivals in Exercise 1 enter quarantine holding under a defined protocol. After seven days, 870 remain alive, 60 were confirmed dead and no additional movements occurred. Conditional survival during holding is 870 divided by 930, or about 93.55 per cent. Cumulative live survival from dispatch through seven days is 870 divided by 1,000, or 87 per cent. These two values have different start points, both useful for different questions. A shipping contractor may focus on arrival; the receiving nursery needs delayed outcomes too. The scientific report should state both, then investigate whether any later deaths plausibly relate to prior transport, local water quality or the general fragility of the developmental stage. Chronology alone is not proof of causation.
Exercise 3: percentages from stages multiply rather than add
Consider a simplified cohort with 90 per cent survival through transport, then 80 per cent conditional survival during nursery holding and 75 per cent conditional survival through a subsequent transfer interval. The cumulative proportion is 0.90 multiplied by 0.80 multiplied by 0.75, or 0.54, equal to 54 per cent of the original cohort. Adding percentages gives a meaningless 245 per cent. Averaging gives about 81.7 per cent but answers neither cumulative nor stage-specific survival. On a starting batch of 2,000, the expected survivors under those illustrative proportions are 1,800 after transport, 1,440 after nursery and 1,080 after transfer. Keep the integers in the ledger where real counts are available rather than rounding each hypothetical stage prematurely. This demonstrates why definitions and arithmetic must remain attached.
Exercise 4: two cages, unequal starting counts
A small nursery cage receives 100 pueruli and produces 90 surviving juveniles. A larger cage receives 900 and produces 630 survivors. Their individual survivals are 90 and 70 per cent. The unweighted cage average is 80 per cent, but the animal-weighted batch survival is 720 of 1,000, or 72 per cent. An operator reporting overall unit output should generally show 72 per cent, with cage-specific outcomes alongside it. A researcher treating cages as experimental replicates may be interested in variation among cage rates, while recognising that cage size and stocking may be confounded. Neither statistic should be presented without explaining its denominator. In particular, the 80 per cent cage-average is not an accurate count of what happened to 1,000 animals.
Exercise 5: extra animals arrive halfway through the trial
A nursery begins with 500 pueruli. By day four, it records 40 confirmed deaths. Then a new consignment of 200 animals is added to the same cage. At day ten, staff count 600 living animals. Dividing 600 by 500 gives a nonsensical 120 per cent survival because the population was open to additions. The balance is 500 plus 200 minus 40 minus 600, leaving 60 animals whose status requires reconciliation if no other movements occurred. Those 60 are not automatically deaths. The better husbandry system keeps cohorts distinct or documents each transfer and uses marked groups or reliable split accounting where appropriate. If cohort identities cannot be resolved, state that a valid initial-cohort survival percentage cannot be reconstructed from the pooled closing inventory.
Exercise 6: a transfer out is not nursery mortality
Six hundred animals were living at a checkpoint. Staff moved one hundred juveniles to another facility with a signed dispatch manifest. The next count found 475 living in the original nursery and 25 confirmed deaths during the interval. The balance is 600 minus 100 dispatched minus 25 dead equals 475 remaining, with no unexplained discrepancy. If an analyst treats all 125 missing from the original cage as deaths, they calculate an erroneous 79.2 per cent survival for the interval. The correct cohort-wide survival, assuming the hundred transferred animals were alive at dispatch and the aim is survival to transfer or continued holding, is 575 of 600, or about 95.8 per cent. Later survival at the receiving facility is a new interval that needs its own count, not an assumption.
Exercise 7: counting an empty shell as a carcass
During a nursery check, a worker finds twenty discarded exoskeletons and enters twenty deaths into the spreadsheet. This error can destroy the apparent survival of an otherwise healthy cohort. Lobsters moult by shedding an exoskeleton, and an empty exuvia is not a carcass. The correction is to record moult evidence in a separate column and count confirmed deceased animals only when that conclusion is justified. Communal conditions can also make it difficult to attribute one shell to one specific individual, so twenty exuviae do not automatically mean twenty different lobsters moulted during the reporting interval. If staff perform a live count, they should use a consistent low-stress method suited to the species and size rather than excessively disturbing shelter to chase an exact total.
Exercise 8: juvenile conversion and survival are not synonyms
A checkpoint begins with 600 living pueruli. The follow-up records 420 juveniles, 120 pueruli still at the transition stage and 60 confirmed dead, with no movements. The survival proportion is 540 divided by 600, or 90 per cent. The juvenile conversion relative to the starting six hundred is 420 divided by 600, or 70 per cent. Among the 540 survivors, the juvenile share is 420 divided by 540, or about 77.8 per cent. All three values are legitimate if labelled correctly; none can replace the others. A report that calls 70 per cent juvenile conversion 70 per cent survival understates the number alive. A report that calls 90 per cent survival a 90 per cent successful metamorphosis rate overstates the number known to have completed the stage transition.
Exercise 9: hiding animals and imperfect detection
A nursery reports sixty animals when the observer scans visible shelter surfaces, then ninety when the enclosure is safely and systematically checked by a trained team. Neither count is automatically the ground truth if animals can remain hidden. The first is a visibility-based count; the second may be a more complete census under its own protocol. Do not declare thirty animals newly born or newly arrived merely because a different counting method detects them. If repeated counts are needed, standardise inspection technique, record counting effort, minimise handling harm and investigate detection probability. A sample estimate should be identified as an estimate, complete with method and uncertainty, not rewritten as an exact physical inventory. Disappearance and death are different categories until there is evidence linking them.
Exercise 10: survival after one day versus survival after one week
Transport approach A delivers 95 living animals from 100 shipped, but only 78 are living after seven days. Approach B delivers 90 from 100 and retains 85 after seven days. At arrival, A appears better, 95 versus 90 per cent. At the shared seven-day checkpoint, B appears better, 85 versus 78 per cent. It would be misleading to claim the second method is superior without considering whether batches, sizes, travel times and rearing conditions were comparable. The important result is that the ranking changes when the follow-up window changes. A complete trial should predefine arrival and delayed outcomes and include both in reporting. The practical decision should not be based only on the more flattering of two time points.
Exercise 11: a temperature maximum without exposure time
A logger shows one measurement at 31 degrees Celsius. The number alone cannot establish thermal injury, because it lacks the sensor location, sampling interval, duration of exposure and species-specific context. A probe in direct sunlight outside a bag could register a different condition from the animals’ water. Conversely, a real prolonged excursion may be important. Reconstruct timestamps, check the logger’s calibration, note whether it was inside a representative transport unit, and compare all measured conditions with a justified husbandry standard. Avoid inventing a universal safe temperature merely to make the case decisive. If mortality later increases, describe the association and alternative stresses, such as handling, crowding and water-quality deterioration. Accurate measurement is the foundation for any causality discussion.
Exercise 12: oxygen addition does not describe the whole transport environment
Two shipments are labelled oxygenated. One used continuous air aeration, while another used oxygen placed into a sealed transport container. The term oxygenated conceals different gas and mixing arrangements. It also gives no information about carbon dioxide accumulation, ammonia, temperature, exposure duration or density. To compare the shipments, reconstruct container design, volume, initial water condition, measured gas readings, temperature history, number of animals and follow-up survival. A high oxygen reading at loading does not establish that conditions stayed adequate for the entire journey. Nor does a missing oxygen measurement prove a shipment was unsafe. Translate the actual recorded procedure, and use a separate quality-assessment field for any conclusion supported by the data.
Exercise 13: first-week losses attributed to cannibalism
A nursery finds fewer animals than the previous count and blames cannibalism because the culture is densely stocked. Cannibalism may be biologically plausible, and shelter research can inform experiments, but a count deficit alone does not establish it. Check for escapes, mesh faults, undocumented removals, counting inconsistency, disease, water incidents and handling damage. Photograph or record relevant observations where appropriate and seek expert diagnosis when disease is suspected. If a controlled shelter comparison is proposed, use replication and comparable starting animal sizes instead of placing all small animals in one treatment and all large animals in another. A responsible interpretation is that losses occurred under the reported conditions and a mechanism is under investigation. That statement directs corrective work without presenting speculation as proof.
Exercise 14: the wrong density denominator
A report states that a cage holds 120 animals per cubic metre, but the calculation actually divides 120 animals by one square metre of bottom surface. Those units are different: 120 per square metre is an areal stocking density, not a volumetric density. In a cage with one square metre of bottom and two cubic metres of water, the volumetric density is 60 per cubic metre. For shelter-using crustaceans, bottom or refuge area may be an important biological variable even where water volume remains relevant to water exchange. Record cage dimensions and the chosen denominator, then avoid comparing an areal figure with a volumetric one. If an experiment specifies one type of density, preserve it in the translation and the worked calculation.
Exercise 15: survival from a subsample is an estimate
A technician samples five of twenty nursery baskets and finds forty-two, thirty-eight, forty-five, forty and thirty-five live animals. The mean is forty per basket; multiplying by twenty yields an estimated 800 survivors across the nursery, not a confirmed count of 800 individuals. This expansion assumes the baskets sampled are representative of the full set. If the five were chosen because they were easiest to reach, the estimate may be biased. Record the basket-selection method, observed range and any size or habitat differences. A confidence interval would need a suitable sampling design and account for variation among baskets. Report approximately 800 estimated living animals under the stated sample design, not exactly 800 animals inspected, and do not convert the difference from an earlier estimate into confirmed deaths.
Exercise 16: when a batch label is reused
A facility reuses the label N-12 for a new consignment once the previous group leaves the nursery. A later analyst merges the two because their spreadsheet IDs match, causing impossible counts and apparently excellent survival. Batch identifiers should be unique across dates and facilities, or combined with a unique creation timestamp and documented source. If historical records are ambiguous, reconcile receiving manifests, enclosure assignments and transfer logs before calculating any cumulative rate. A repaired database can preserve the original label as a displayed field while assigning a true unique key behind it. Biology cannot be extracted reliably from a ledger that combines animals that were never in the same cohort. Stable identity is not bureaucratic decoration; it is part of measurement.
Exercise 17: provenance and the hatchery label
A dealer sells 2,000 wild-caught pueruli to a nursery, which successfully raises many juveniles. An advertisement then calls the juveniles hatchery-born. That language is inaccurate if the animals were actually collected from the wild and only nursed under controlled conditions. A true closed-cycle hatchery claim should be backed by parental or egg origin, phyllosoma rearing and documented stage progression. It is possible to describe the legitimate value of nursery husbandry without claiming a life cycle that did not occur. Provenance is also relevant to licensed collection, stock management and disease-risk controls. A reader should be able to trace each lot to an authorised source, not merely to the most recent farm that handled it.
Exercise 18: quality of evidence for a husbandry trial
Suppose method A is used on one cage with 400 animals during a cool week, while method B is used on a second cage with 400 during a warm week. Method A produces 320 survivors, or 80 per cent, and B produces 280, or 70 per cent. The ten-percentage-point contrast is observed, but time, environmental conditions and cage identity also differ. There is no replicated comparison that separates those factors. A stronger test allocates comparable animals among replicated enclosures and controls or records relevant conditions. Reports should provide sample sizes and the experimental unit, not just percentages. A beautiful chart can be statistically weak when each treatment has only one true replicate.
Exercise 19: fixed starting inventory but shifting definitions
A nursery begins a trial with 900 accepted living animals. One weekly report uses that number as its denominator, while the next excludes 50 animals judged initially weak and reports survival from 850. Both ratios may be useful for their stated subgroups, but mixing them creates a false trend. Keep the all-accepted cohort intact and report subgroup outcomes separately, with consistent inclusion rules. If a retrospective exclusion is necessary, document the rationale and show sensitivity of conclusions to the choice. Selecting only robust animals after seeing which survived creates selection bias. A reader should be able to tell whether a comparison is about real improvement or an altered definition of who counted at the start.
Exercise 20: transfer-ready is a decision, not simply a count
At the end of a nursery trial, 700 lobsters are alive, but only 560 meet an operational transfer-readiness definition based on documented developmental stage, condition and size. The live survival fraction from an initial 1,000 is 70 per cent. Transfer-ready yield is 560 divided by 1,000, or 56 per cent, and the share of survivors meeting the readiness definition is 560 divided by 700, or 80 per cent. These numbers serve different management questions. Clearly define the readiness criteria and do not invent a universal cutoff across species or farms. A high live count may still lead to delayed production if many animals are not yet ready for the next phase. Operational decisions need the right outcome, not merely a high percentage.
An independent nursery calculation readers can carry to another species
Given an opening cohort, write the inventory equation before writing a survival percentage. Check what came in, what left, what died, what remained and what cannot be accounted for. Define the clock for each stage and separate observed stage transitions from survival. When pooled cages differ in initial size, show animal-weighted totals and cage-level rates. When measurements are estimates rather than full counts, label them accordingly. This method is transferable to other hatchery and nursery records because it is fundamentally a way of keeping the units, identity and observation window honest.
The most valuable sentence in a technical report is often the one that acknowledges what was not established. A low measured survival rate identifies a problem worth fixing, but it does not prove disease, cannibalism or transport damage on its own. A high rate demonstrates performance for a defined cohort, but it cannot guarantee the next shipment. A complete interpretation preserves these boundaries while still giving managers the concrete count reconciliation needed to improve practice.
Related eduKate explanations
Spiny lobster puerulus collectors and settlement indices explains the earlier detection stage. Abalone juvenile nursery density and transfer survival explores comparable denominator problems in another animal. Sandfish seed-hatchery unit costs and survival illustrates why early survival influences later unit costs.
Sources and further reading
ACIAR, Research for development of lobster growout technology in Indonesia. First-party project history and objectives.
James Cook University repository, final report of the Indonesia lobster growout technology project. Details of nursery, early juvenile mortality and nutrition research.
ACIAR, Spiny lobster aquaculture in the Asia-Pacific region. Specialist proceedings covering nursery and grow-out.
Global Seafood Alliance, Spiny lobster aquaculture. Historical synthesis of the life cycle and nursery methods.
FAO AGRIS, Effects of shelter types and feeding rates in juvenile Panulirus ornatus. Research record of an experimental shelter and ration study.
All numerical scenarios in this guide are fictional teaching examples, not actual company or research-farm records.