Fish Stock Assessment Methods
Expert-defined terms from the Professional Certificate in Fish Stock Assessment Economics course at London School of Business and Administration. Free to read, free to share, paired with a professional course.
A – Age‑structured model #
A quantitative framework that groups fish by age classes to estimate mortality, growth, and recruitment.
Example #
Using catch‑at‑age data to project future stock size.
Application #
Supports management decisions on harvest limits.
Challenge #
Requires accurate age determination and long‑term data series.
A – Acoustic survey #
A method that employs sound waves to estimate fish abundance and distribution.
Example #
Deploying a vessel‑mounted echosounder to map sardine schools.
Application #
Provides rapid, non‑invasive abundance indices.
Challenge #
Calibration to species‑specific target strength is essential.
A – Absolute abundance #
The total number of individuals in a population, not relative to any reference.
Example #
Estimating total cod numbers from a trawl survey.
Application #
Basis for reference points such as MSY.
Challenge #
Often difficult to obtain without exhaustive sampling.
A – Annual catch #
The total weight or number of fish harvested in a calendar year.
Example #
120 000 t of tuna reported in 2023.
Application #
Input for stock assessment models.
Challenge #
Under‑reporting and illegal catches can bias estimates.
B – Biomass #
The total weight of all individuals in a stock, commonly expressed in metric tonnes.
Example #
Estimated spawning biomass of herring at 2 Mt.
Application #
Used to calculate reference points like BMSY.
Challenge #
Converting abundance indices to weight requires reliable length‑weight relationships.
B – Biomass dynamics model #
A model that predicts changes in stock biomass over time based on growth, recruitment, and mortality.
Example #
Applying the Schaefer model to a small pelagic fishery.
Application #
Provides simple forecasts for data‑limited stocks.
Challenge #
Assumes constant productivity, which may not hold under environmental change.
B – Biological reference point #
Benchmarks such as MSY, BMSY, and FMSY that guide sustainable harvest levels.
Example #
Setting the target fishing mortality at 0.3 yr⁻¹ for a demersal species.
Application #
Aligns fisheries management with conservation objectives.
Challenge #
Estimating reference points accurately requires robust stock assessments.
C – Catch‑at‑age data #
Records of fish caught, categorized by age, used to reconstruct population dynamics.
Example #
Analyzing the proportion of 3‑year‑old cod in the catch.
Application #
Informs age‑structured assessment models.
Challenge #
Age misclassification can lead to biased mortality estimates.
C – Catch‑per‑unit‑effort (CPUE) #
A measure of catch relative to fishing effort, serving as an index of abundance.
Example #
0.5 kg hook⁻¹ for a longline fishery.
Application #
Provides temporal trends in stock abundance.
Challenge #
CPUE may not be proportional to true abundance due to hyperstability or hyperdepletion.
C – Coastal transect survey #
A systematic sampling method that follows a predetermined line along the coast to assess fish communities.
Example #
Divers recording reef fish counts along a 5 km transect.
Application #
Generates spatially explicit abundance data.
Challenge #
Weather conditions and diver bias can affect data quality.
C – Component stock assessment #
An assessment focusing on a single species or stock within a mixed‑species fishery.
Example #
Assessing the Atlantic mackerel stock separately from other pelagics.
Application #
Allows targeted management actions.
Challenge #
Interactions with other species may be overlooked.
D – Data‑limited assessment #
An approach used when few or no fishery‑dependent data are available, often relying on catch histories or ecological proxies.
Example #
Using the CMSY method for a newly exploited species.
Application #
Provides preliminary status for emerging fisheries.
Challenge #
High uncertainty and sensitivity to assumptions.
D – Depletion #
The proportion of a stock’s current biomass relative to its unfished biomass (B/B₀).
Example #
A depletion level of 0.4 indicates 60 % reduction from unfished levels.
Application #
Helps determine if a stock is overfished.
Challenge #
Estimating unfished biomass accurately is often problematic.
D – Depth stratified sampling #
Collecting fish samples at multiple depth intervals to capture vertical distribution patterns.
Example #
Conducting trawls at 50 m, 100 m, and 150 m.
Application #
Improves representativeness of abundance estimates.
Challenge #
Requires precise depth control and increased vessel time.
E – Exploitation rate (U) #
The proportion of the stock that is removed by fishing each year, often expressed as F/M.
Example #
U = 0.3 yr⁻¹ indicates 30 % of the stock harvested annually.
Application #
Guides setting of catch limits.
Challenge #
Estimating U accurately depends on reliable mortality and catch data.
E – Exploitation status #
Classification of a stock (e.g., overfished, fully exploited, underfished) based on its depletion and exploitation rate.
Example #
A stock with B/B₀ = 0.35 and U > U_MSY is deemed overfished.
Application #
Determines need for recovery measures.
Challenge #
Thresholds may vary among jurisdictions.
F – Fisheries independent survey #
Scientific sampling that is not influenced by commercial fishing effort, providing unbiased abundance indices.
Example #
A NOAA bottom‑trawl survey of groundfish.
Application #
Supplies data for assessment models.
Challenge #
Expensive and logistically demanding.
F – Fisheries dependent data #
Information derived from commercial or recreational fishing activities, such as catch, effort, and size composition.
Example #
Vessel‑monitoring system records of daily landings.
Application #
Forms the backbone of many assessment models.
Challenge #
May suffer from reporting bias and data gaps.
F – Fishing mortality (F) #
The rate at which fish are removed by fishing, expressed per unit time.
Example #
F = 0.45 yr⁻¹ for a heavily fished demersal stock.
Application #
Central parameter in age‑structured models.
Challenge #
Separating F from natural mortality (M) can be difficult.
F – Fisheries management strategy evaluation (MSE) #
A simulation framework that tests the performance of management procedures under uncertainty.
Example #
Evaluating a harvest control rule for a tuna fishery.
Application #
Helps design robust policies before implementation.
Challenge #
Requires extensive computational resources and expert knowledge.
G – Growth parameters (von Bertalanffy) #
Constants (L∞, k, t₀) describing how fish length increases with age.
Example #
L∞ = 120 cm, k = 0.2 yr⁻¹ for a cod population.
Application #
Converts age data to length and weight for biomass estimates.
Challenge #
Parameter values may vary spatially and temporally.
G – Gross stock productivity #
The maximum sustainable yield (MSY) expressed as a proportion of unfished biomass (MSY/B₀).
Example #
A productivity of 0.3 t yr⁻¹ t⁻¹ indicates moderate growth potential.
Application #
Informs selection of appropriate assessment models.
Challenge #
Highly sensitive to assumptions about recruitment.
H – Harvest control rule (HCR) #
A pre‑agreed algorithm that translates stock status into allowable catch or effort.
Example #
Reducing TAC by 10 % for every 0.05 drop in B/B_MSY.
Application #
Provides transparent, repeatable management actions.
Challenge #
Designing HCRs that are both precautionary and flexible.
H – Historical catch reconstruction #
The process of estimating past catches using indirect evidence, such as trade data or anecdotal reports.
Example #
Adding unreported artisanal catches to official statistics.
Application #
Improves long‑term trend analyses.
Challenge #
Uncertainty can be large, especially for early periods.
I – Index of abundance #
A metric derived from surveys that reflects relative stock size, often standardized for variation in gear or conditions.
Example #
Annual mean trawl catch per unit effort for haddock.
Application #
Input for stock assessment models.
Challenge #
Requires rigorous standardisation to avoid bias.
I – Integrated assessment #
Combining multiple data sources (e.g., fisheries‑dependent, independent, biological) within a single modelling framework.
Example #
Merging acoustic survey biomass with age‑structured catch data.
Application #
Increases robustness of stock status estimates.
Challenge #
Complex statistical methods and data compatibility issues.
J – Joint likelihood #
A statistical approach that simultaneously fits several data sets to a common set of model parameters.
Example #
Combining catch, CPUE, and length composition in a single likelihood function.
Application #
Improves precision of parameter estimates.
Challenge #
Requires careful weighting of heterogeneous data.
K – K‑selectivity #
The relationship between gear selectivity and fish length or age, often described by a logistic function.
Example #
50 % retention at 30 cm for a gillnet.
Application #
Essential for age‑structured models to account for size‑dependent capture.
Challenge #
Estimating selectivity without bias from gear changes.
L – Length‑frequency data #
Distribution of fish sizes in a sample, used to infer growth, mortality, and recruitment patterns.
Example #
Histogram of 5‑cm length bins for a snapper catch.
Application #
Basis for length‑based assessment methods.
Challenge #
Requires conversion to age with growth parameters; susceptible to sampling bias.
L – Length‑based spawning potential ratio (LB‑SPR) #
A metric estimating the proportion of mature biomass relative to unfished conditions using length data.
Example #
LB‑SPR of 0.45 indicating reduced reproductive capacity.
Application #
Useful for data‑limited stocks where age data are unavailable.
Challenge #
Relies on accurate maturity ogives and length‑weight relationships.
M – Mark‑recapture study #
A method where individuals are captured, marked, released, and later recaptured to estimate population size.
Example #
PIT‑tagging 1 000 cod and recapturing 150 after one year.
Application #
Provides direct estimates of abundance and mortality.
Challenge #
Requires sufficient recaptures and assumptions of closed populations.
M – Maximum sustainable yield (MSY) #
The largest average catch that can be taken indefinitely without depleting the stock.
Example #
MSY set at 45 000 t for a pelagic fishery.
Application #
Serves as a target for many management regimes.
Challenge #
Uncertainty in stock dynamics may render MSY unattainable or unsafe.
M – Mean length at age #
Average fish length for a given age cohort, derived from growth models or empirical data.
Example #
Mean length of 2‑year‑old herring is 18 cm.
Application #
Transforms age data into weight for biomass calculations.
Challenge #
High variability due to environmental effects.
N – Natural mortality (M) #
The rate at which fish die from non‑fishing causes (predation, disease, senescence).
Example #
M estimated at 0.2 yr⁻¹ for a fast‑growing species.
Application #
Required to separate fishing impacts from natural processes.
Challenge #
Direct measurement is rare; often inferred from empirical formulas.
N – Number‑at‑age (Nₐ) #
The count of individuals in each age class at a specific time, fundamental to age‑structured models.
Example #
N₁₀ = 2 × 10⁶ for ten‑year‑old cod.
Application #
Drives calculation of spawning biomass and catch.
Challenge #
Sensitive to errors in age determination and selectivity.
O – Operating model #
A simulated representation of a fishery and its ecosystem used in MSE to test management procedures.
Example #
An operating model that includes recruitment variability and fleet dynamics.
Application #
Provides realistic scenarios for evaluating HCR performance.
Challenge #
Must capture key sources of uncertainty without being overly complex.
O – Overfished status #
A condition where the stock biomass is below the threshold defined by a reference point (e.g., B < 0.5 B_MSY).
Example #
A stock with B/B_MSY = 0.3 is classified as overfished.
Application #
Triggers recovery measures such as catch reductions.
Challenge #
Determining accurate biomass estimates to avoid misclassification.
P – Population dynamics model #
A mathematical representation of the processes governing stock size, including growth, recruitment, and mortality.
Example #
The Ricker stock‑recruitment model applied to a sardine fishery.
Application #
Generates forecasts under different management scenarios.
Challenge #
Model choice influences outcomes; parameter uncertainty can be high.
P – Production model #
A type of surplus‑production model that relates catch to changes in biomass, often assuming logistic growth.
Example #
Estimating B₀ and r for a small‑scale fishery using the Schaefer model.
Application #
Provides quick assessments when data are scarce.
Challenge #
Assumes a constant carrying capacity, which may not hold under environmental shifts.
Q – Quota allocation #
Distribution of total allowable catch among fishing entities or sectors.
Example #
Assigning 30 % of the haddock quota to community fisheries.
Application #
Controls effort distribution and supports socio‑economic objectives.
Challenge #
Allocation must balance equity, compliance, and biological sustainability.
R – Recruitment #
The addition of new individuals to the fishable population, typically measured as age‑1 fish entering the fishery.
Example #
Year‑class of 800 000 t of anchovy recruits in 2022.
Application #
Central driver of stock productivity.
Challenge #
Highly variable and influenced by environmental factors.
R – Recruitment‑spawning stock biomass (R‑SSB) relationship #
The functional link describing how spawning biomass influences recruitment.
Example #
A Beverton‑Holt curve with steepness parameter h = 0.75.
Application #
Predicts future stock size under different exploitation levels.
Challenge #
Parameter estimation is often uncertain due to limited data.
S – Spawning biomass (SB) #
The total weight of mature individuals capable of reproduction, a key indicator of reproductive capacity.
Example #
SB = 1.2 Mt for a northern cod stock.
Application #
Used to set B_MSY and assess overfishing.
Challenge #
Requires accurate maturity schedules and length‑weight conversions.
S – Spawning potential ratio (SPR) #
Ratio of spawning biomass under current fishing mortality to that under unfished conditions.
Example #
SPR = 0.4 indicating 60 % reduction from virgin spawning capacity.
Application #
Guides setting of F targets (e.g., SPR = 0.35).
Challenge #
Sensitive to assumptions about selectivity and maturity.
S – Standardised CPUE #
CPUE data that have been statistically adjusted for variations in gear, location, and other covariates.
Example #
Using a generalized additive model to remove seasonal effects.
Application #
Produces comparable abundance trends over time.
Challenge #
Model misspecification can introduce bias.
T – Time‑varying parameter #
Model parameters that are allowed to change across years to reflect environmental or management influences.
Example #
Allowing recruitment deviation to follow a random walk.
Application #
Captures non‑stationarity in stock dynamics.
Challenge #
Increases model complexity and requires more data.
T – Truncated catch curve #
A method for estimating total mortality by fitting a linear regression to the descending part of the log‑catch‑at‑age plot, excluding the youngest ages.
Example #
Using ages 3–7 to estimate Z in a cod stock.
Application #
Provides quick mortality estimates.
Challenge #
Choice of truncation point can affect results.
U – Unfished biomass (B₀) #
The theoretical biomass of a stock in the absence of fishing, serving as a reference for depletion calculations.
Example #
B₀ estimated at 3 Mt for a pelagic species.
Application #
Basis for computing B/B₀ and SPR.
Challenge #
Often inferred from models rather than observed directly.
V – Vulnerability #
The susceptibility of a stock to fishing pressure, encompassing selectivity, behavior, and habitat exposure.
Example #
High vulnerability of surface‑dwelling sardines to purse‑seine gear.
Application #
Informs gear regulations and spatial closures.
Challenge #
Quantifying vulnerability requires multidisciplinary data.
W – Weight‑at‑age #
The average weight of fish in each age class, derived from length‑weight relationships.
Example #
2‑year‑old cod weigh on average 1.5 kg.
Application #
Converts age composition to biomass estimates.
Challenge #
Intraspecific variability can cause errors.
Y – Yield‑per‑recruit (Y/R) model #
A calculation that predicts the average catch per recruit as a function of fishing mortality and life‑history parameters.
Example #
Y/R curve showing maximum yield at F = 0.35 yr⁻¹.
Application #
Assists in selecting optimal fishing mortality.
Challenge #
Assumes constant recruitment and may not capture ecosystem interactions.
Z – Total mortality (Z) #
The sum of fishing mortality (F) and natural mortality (M), representing the overall rate at which fish leave the population.
Example #
Z = 0.6 yr⁻¹ for a heavily fished stock (F = 0.4, M = 0.2).
Application #
Central to age‑structured assessments and catch‑curve analyses.
Challenge #
Accurate partitioning of Z into F and M is often uncertain.