Age-structured Stock Assessment Research Articles

The robustness of our assumptions about recruitment: A re-examination of marine recruitment dynamics with additional data and novel methods

Recruitment is the number of young fish entering a population and is an essential process in age-structured stock assessment models. Many of the models on which stock assessments are based assume some level of influence of spawning biomass on recruitment, and estimation of this functional relationship has been a focus of substantial research. However, recruitment is also influenced by environmental fluctuations that induce autocorrelated patterns, trends, and shifts in deviations from the stock-recruitment relationship. Consequently, previous meta-analyses of stock-recruitment relationships have had trouble calibrating the relationship. We revisited the findings of Szuwalski et al. (2015) to determine if their results are robust to the addition of data to the RAM Legacy Stock Assessment database and choice of analysis method (the PELT algorithm, MARSS models, and Bayesian change point detection). Since the publication of Szuwalski et al. (2015), the number of stocks in the database has doubled. We determined the primary influence of spawning biomass on recruitment and examined the recruitment time series for regime shifts for 432 stocks. Our results indicated that 57% of stocks did not have a significant correlation between spawning biomass and recruitment over the observed biomasses. Environmental conditions played a larger role in recruitment variation than spawning biomass. The presence, location, and number of regime shifts in recruitment time series was highly dependent on the detection method, with Bayesian change point detection identifying the fewest regime shifts. Despite the sensitivity to the detection method, 46% of stocks without a significant correlation between spawning biomass and recruitment are estimated to have experienced at least one regime shift as determined by the PELT algorithm Our analyses suggest that effective methods for modeling and forecasting large variations in recruitment over time are needed, particularly given that climate change is predicted to impact the frequency and magnitude of regime shifts.

Spatial estimation and yearly trends in abundance-index of Japanese jack mackerel (Trachurus japonicus) in the East China Sea and Sea of Japan

Catch per unit effort (CPUE) of commercial fisheries is recommended to be standardized because various factors can affect catch rates. Recently, the spatio-temporal analysis has been required to standardize fishery data, because catch is spatially distributed in proportion to fishery distribution and not in proportion to fish abundance. We developed the spatio-temporal models for the CPUE of large- and medium-sized purse seine fisheries for the different age groups of Japanese jack mackerel (Trachurus japonicus) to evaluate the potential spatial distribution and estimate the annual trends in abundance index. Japanese jack mackerel is widely distributed in the East Asian seas, including the East China Sea and Sea of Japan, and is an important fishery resource in Japan. Our models were delta-lognormal generalized additive mixed models that incorporated the temporal components of year and month, and size of fishing vessels as categorical variables, and the identity of the fishing vessels as a random variable. The spatial component represented by the latitude and longitude points of the fishing location was smoothed using an isotropic bivariate function. The water temperature at a depth of 50 m was incorporated to detect the effects of environmental factors. After all combinations of explanatory variables were evaluated, the model with the lowest Bayesian information criterion score was selected as the best model. All the best models among both binomial models and lognormal models contained all explanatory variables, although the best binomial model for age 2 exclude the vessel size and the best lognormal model for age 0 exclude the water temperature at a depth of 50 m. Our results indicate that the potential distribution areas for all ages of Japanese jack mackerel exist on the continental shelf in the southern part of the East China Sea, the western side of Kyushu Island, and along Honshu Island in the western part of the Sea of Japan. In contrast, the main distribution areas with higher density for age 0 were segregated into two: the southwestern East China Sea and the coastal area off Kyushu Island, although fish over age 1 were widely distributed. The standardized CPUE was higher than the nominal CPUE for age 0 from 2015 to 2020, and the standardized CPUE was also higher for age 1 in 2020. These results were derived from spatial models that compensated for the lack of spatial data due to the recent decrease in fishing operations in the Southeast China Sea. The yearly trends of standardized CPUE for each age could be used for the age-structured stock assessment model such as a tuned virtual population analysis. Improving the accuracy of the abundance index helps to further improve the reliability of stock assessments.

Analytical expressions of sustainable benchmarks generic for all two-stage-structured models of exploited fish populations

Of the types of fishery models that use time series of removals, fishing effort, and indices of abundance, two-stage-structured models are less prevalent, especially in developing the maximum sustainable yield (MSY) benchmarks, in part because of ignorance of how to calculate those benchmarks. To fill this gap, this study aims to derive analytical expressions for calculating MSY benchmarks associated with a continuous delay-differential model, given that the recruit stage relates to the spawning-stock biomass (SSB) according to a stock–recruit relationship (SRR) and that the SRR parameters are combined with the components of the composite (two-stage-structured or age-aggregated) yield per-recruit model CYPR14. The focus is on analytical expressions of the fishing mortality producing MSY ( FMSY), considering that the equilibrium yield comes from either the equilibrium SSB at the start of fishing regimes or the equilibrium average SSB during fishing regimes. The expressions obtained are versatile for all two-stage-structured models. Derivations employing the fished equilibrium average SSB are recommended for use because they produce precautionary MSY benchmarks and because they are consistent with their counterparts in contemporary age-structured stock assessment models.

Natural mortality in exploited fish stocks: annual variation estimated with data from trawl surveys

Abstract Natural mortality (M) is difficult to measure directly in exploited fish stocks. Therefore, in stock assessments, M has usually been assumed to be constant for age classes of fishable size. However, M is likely to vary annually due to temporal changes in food availability, predation, fish size at age, and other biological and environmental factors. We suggest that M of fish at age and year is the sum of length-related mortality and condition-related mortality (M = ML + MC). We propose that MC is directly proportional to PKH, the percentage of fish in the stock with condition factor (K) and liver condition (H) below critical values. This relationship is used to demonstrate likely patterns in annual variation in MC for ages 2–9+ in Icelandic cod (Gadus morhua) based on fish measurements in groundfish surveys in spring and autumn. There were annual fluctuations and temporal trends observed in MC for different age classes attributable to variation in food availability. The described method is suitable for age-structured stock assessments of cod and other fish stocks. Our results indicate that basic stock assessments may be improved by including estimates of variable MC.

Ocean model-based covariates improve a marine fish stock assessment when observations are limited

Abstract The productivity of many fish populations is influenced by the environment, but developing environment-linked stock assessments remain challenging and current management of most commercial species assumes that stock productivity is time-invariant. In the Northeast United States, previous studies suggest that the recruitment of Southern New England-Mid Atlantic yellowtail flounder is closely related to the strength of the Cold Pool, a seasonally formed cold water mass on the continental shelf. Here, we developed three new indices that enhance the characterization of Cold Pool interannual variations using bottom temperature from a regional hindcast ocean model and a global ocean data assimilated hindcast. We associated these new indices to yellowtail flounder recruitment in a state–space, age-structured stock assessment framework using the Woods Hole Assessment Model. We demonstrate that incorporating Cold Pool effects on yellowtail flounder recruitment reduces the retrospective patterns and may improve the predictive skill of recruitment and, to a lesser extent, spawning stock biomass. We also show that the performance of the assessment models that incorporated ocean model-based indices is improved compared to the model using only the observation-based index. Instead of relying on limited subsurface observations, using validated ocean model products as environmental covariates in stock assessments may both improve predictions and facilitate operationalization.

A comparison of 4 primary age-structured stock assessment models used in the United States

A comparison of 4 primary age-structured stock assessment models used in the United States

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. material

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. material

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. figs. 1-2

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. figs. 1-2

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. table 3

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. table 3

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. table 1

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. table 1

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. table 2

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. table 2

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. figs. 3-8

A comparison of 4 primary age-structured stock assessment models used in the United States: suppl. figs. 3-8

Effect of food availability on the growth and age determination of European sardine (Sardina pilchardus Walbaum, 1792) larvae

AbstractAccurate assessment of age and growth of fish is essential to understand population dynamics, namely for age-structured stock assessment and for determining vital rates of fish (e.g. age at sexual maturity) and their relationship with environmental variability. To validate the daily deposition in the otoliths of European sardine (Sardina pilchardus) larvae, these were reared with a range of food densities translating into feeding rates extending from less than required for maintenance to saturated feeding levels. When exogenous feeding began, a high variability in the relationship between number of growth increments (GI) and elapsed days after hatching (dph) was observed, irrespective of the feeding treatment. GI counts using a light microscope were less than one per day for larvae <15 dph (ranging from 0.17–0.43 GI day−1) and similar for larvae reared with different food concentrations. The rate of GI count vs age was significantly higher for larvae older than 15 dph. GI count from 3–30 dph was lower than one per day (0.45–0.75 GI day−1 95% CI) for fed larvae 3–30 dph. Scanning electron microscopy confirmed the daily deposition of GI and revealed some increment widths around 0.2 μm, particularly during the first week post-hatch. Otoliths of wild sardine larvae exhibited narrow GI (<0.5 μm) comparable with reared larvae. This study demonstrates that during the first weeks after hatch the daily increment deposition is underestimated using traditional light microscopy, which must be taken into account in future works determining wild sardine larval growth.

“The Elephant in the Room”: Exploring Natural Mortality Uncertainty in Statistical Catch at Age Models

The natural mortality rate (M) of a fish stock is typically highly influential on the outcome of age-structured stock assessment models, but at the same time extremely difficult to estimate. In data-limited stock assessments, M usually relies on a range of empirically or theoretically derived M estimates, which can vary vastly. This article aims at evaluating the impact of this variability in M using seven Mediterranean stocks as case studies of statistical catch-at-age assessments for information-limited fisheries. The two main bodies carrying out stock assessments in the Mediterranean and Black Seas are European Union’s Scientific Technical Economic Committee for Fisheries (STECF) and Food and Agriculture Organization’s General Fisheries Commission for the Mediterranean (GFCM). Current advice in terms of fishing mortality levels is based on a single “best” M assumption which is agreed by stock assessment expert working groups, but uncertainty about M is not taken into consideration. Our results demonstrate that not accounting for the uncertainty surrounding M during the assessment process can lead to strong underestimation or overestimation of fishing mortality, potentially biasing the management process. We recommend carrying out relevant sensitivity analyses to improve stock assessment and fisheries management in data-limited areas such as the Mediterranean basin.

Ecological Reference Points for Atlantic Menhaden Established Using an Ecosystem Model of Intermediate Complexity

Atlantic menhaden (Brevoortia tyrannus) are an important forage fish for many predators, and they also support the largest commercial fishery by weight on the U.S. East Coast. Menhaden management has been working toward ecological reference points (ERPs) that account for menhaden’s role in the ecosystem. The goal of this work was to develop menhaden ERPs using ecosystem models. An existing Ecopath with Ecosim model of the Northwest Atlantic Continental Shelf (NWACS) was reduced in complexity from 61 to 17 species/functional groups. The new NWACS model of intermediate complexity for ecosystems (NWACS-MICE) serves to link the dynamics of menhaden with key managed predators. Striped bass (Morone saxatilis) were determined to be most sensitive to menhaden harvest and therefore served as an indicator of ecosystem impacts. ERPs were based on the tradeoff relationship between the equilibrium biomass of striped bass and menhaden fishing mortality (F). The ERPs were defined as the menhaden F rates that maintain striped bass at their biomass target and threshold when striped bass are fished at their Ftarget, and all other modeled species were fished at status quo levels. These correspond to an ERP Ftarget of 0.19 and an ERP Fthreshold of 0.57, which are lower than the single species reference points by 30–40%, but higher than current (2017) menhaden F. The ERPs were then fed back into the age-structured stock assessment model projections to provide information on total allowable catch. The ERPs developed in this study were adopted by the Atlantic menhaden Management Board, marking a shift toward ecosystem-based fishery management for this economically and ecologically important species.

How fisher behavior can bias stock assessment: insights from an agent-based modeling approach

During stock assessment, fishery-dependent observations are often used to develop indices of abundance or biomass from catch per unit of effort (CPUE) and contribute catch at size or age information. However, fisher behavior, rather than scientific sampling protocols, determines the spatial and temporal locations of fishery-dependent observations. As a result, trends from fishery-dependent data may be a function of fishing activity rather than fish population changes. This study evaluates whether data collected from commercial fishing fleets in the Gulf of Mexico are representative of trends in fish population size. A coupled bioeconomic agent-based model was developed to generate simulated fishery data, which were used to populate an age-structured stock assessment. Comparison of stock assessment results with simulated fish population dynamics showed that management advice from assessment models based on fishery-dependent data could be biased. Assessment of fish with small home ranges harvested by fishing fleets that frequent the same fishing grounds could cause overestimation of fishing mortality. Not accounting for the spatial structure of the fishers or fish can cause biased estimates of population status.

Age estimation, corroboration and back-calculation of four-spot megrim (Lepidorhombus boscii) on the Porcupine Bank using otoliths

Age and growth of four-spot megrim (Lepidorhombus boscii) at the Porcupine Bank, an important European demersal fisheries area, was estimated for the first time based on the analysis of 1334 otoliths. The state of the stock in the Celtic Seas (including the Porcupine Bank) and the northern Bay of Biscay (Div. 7.b-k and 8.a,b,d) has been recently assessed in ICES mainly based on survey abundance indices, but biological information for a forthcoming age-structured stock assessment process is required. In this study, age was estimated by counting the translucent bands (annuli) of the sagittal otoliths following internationally standardized protocols. Otoliths were obtained from specimens caught during five annual bottom trawl surveys (2008–2012) carried out at the Porcupine Bank. Consistency in age interpretation of four-spot megrim was demonstrated by analyzing the regularity of the distances of the annuli to the otolith's primordium and by the back-calculation of the annuli. The growth patterns obtained from direct otolith age estimation and from back-calculation were also compared. Age estimation was corroborated by tracking year-class abundance indices from the surveys. The strength of the year-classes was well tracked in abundant first age groups, highlighting the very abundant 2008 year-class. Furthermore, the von Bertalanffy growth parameters by sex and for combined sexes were also estimated and compared with those from other stocks. The results obtained for these parameters in the combined sexes direct age estimation, their back-calculation and the combination of both methods (L∞: 50.093; k: 0.126; t0: −0.296) will allow the upcoming age-structured assessment of this stock and thus a suitable population management. The strength of the cohorts here obtained was compared with those from a previous study in the Iberian Atlantic (8.c, 9.a) stock and based on age estimates calculated with the same ageing criterion. The differences in demographic structure and cohort dynamics found between stocks support the requirement that each stock continues to be assessed based on its respective age estimates and length distributions.

The unknown Baranov. Forty years of polemics over the formal theory of the life of fishes

Abstract The year 2018 marked the 100th anniversary of the publication of paper “On the question of the biological basis of fisheries” by F.I. Baranov considered a cornerstone paper of modern fishery science. Baranov formalized population dynamics by describing changes in population abundance using differential equations, introducing the concept of instantaneous fishing and natural mortality rates, and developing his catch equation, which is the foundation of most modern age-structured stock assessment models. Baranov was the first to show the effect of fishing on population structure based on theoretical grounds. At the time of its publication, Baranov’s paper did not receive much attention in Russia and was completely unknown to scientists in the West. The second publication (On the question of the dynamics of the fishing industry, 1925) received substantial criticism from many and sparked a furious debate between Baranov and his opponents that lasted for several decades. The history and content of those debates, expressed in multiple papers by Baranov, is still largely unknown. I describe the essence of arguments by Baranov and his opponents. The story of these scientific debates reveals how different philosophical concepts and dominant points of view evolved through time.

Influence of Growth and Recruitment Parameters in the Assessment and Management Variables of the Yellow Squat Lobster (Cervimunida johni)

Fitting length data in age-structured stock assessment is a common method for evaluating hard-to-age animals, such as crustaceans. Growth specification and the uncertainty in the stock recruitment relationship are key issues in length-based assessment models. We conducted sensitivity analyses to evaluate the impact of growth and recruitment parameters on the stock assessment and management variables of the yellow squat lobster (Cervimunida johni) caught off the Chilean coast. Nine different scenarios of the length at first capture ( L a = 1 ) and the coefficient of variation at age ( c v a ) were tested for six combinations of values for the steepness parameter (h) and the recruitment variance ( σ R 2 ). We also investigated the reliability of these estimates using an operating model. Our findings indicate that the parameter related to growth, L a = 1 , has the greatest impact on the assessment and management variables of this fishery resource, with c v a having a lesser effect. Recruitment and fishing mortality estimates were the main variables affected. Parameters h and σ R 2 did not profoundly impact the variables assessed. In addition, L a = 1 was the most biased estimated parameter. We discuss that the high influence of growth parameters is related to model structure, and thus implications for determination of the status of yellow squat lobster should be addressed in the future. We recommended developing simulation protocols for the selection of growth parameters when using an age-structured model with length observations, and we believe that our findings are relevant for all Chilean fisheries with a similar stock assessment framework.

Dynamics of lake trout production in the main basin of Lake Huron

Abstract To inform lake trout (Salvelinus namaycush) fishery management in Lake Huron that has undergone rapid ecosystem changes, we quantified lake trout production dynamics by coupling age-structured stock assessment and fish bioenergetics models. Our approach revealed the connection between piscivore production and prey consumption, included growth compensation to reproduction losses, and allowed comparisons between long-term dynamics of fishery harvests and fish production. We found that, despite the collapse of alewives, a major non-native pelagic prey fish, lake trout production appeared to be sustainable. To a certain degree, the effect of recent recruitment declines on lake trout production was offset by the release of harvest pressure from subadult lake trout and reduction in fishing and sea lamprey-induced mortality on adult lake trout. Evidence for sustainability also included the finding that no changes in average ratios of annual production to beginning-of-the-year biomass. Juvenile P:B ratio remained as high as 2.1. The effect of growth declines on adult and subadult production was offset by reduction in population mortality. Body growth and condition did not continue to decline when lake trout became more and more reliant on round goby as food, and the dynamics of total consumption of prey fish continued to be recipient controlled.