Estimation of age composition from length data by posterior probabilities based on a previous growth curve: application to Sebastes schlegelii

2005 ◽  
Vol 62 (11) ◽  
pp. 2475-2483 ◽  
Author(s):  
Katsuhisa Baba ◽  
Masayoshi Sasaki ◽  
Noriyuki Mitsutani
Keyword(s):  
Growth Curve ◽  
Estimation Error ◽  
Fish Population ◽  
Fish Growth ◽  
Likelihood Method ◽  
Length Variation ◽  
Posterior Probabilities ◽  
Age Composition ◽  
Sebastes Schlegelii ◽  
Length Data

We developed a system to estimate the age composition of a fish population (Sebastes schlegelii) from length data by considering fish growth, length variation, proportion of age classes, and sexual dimorphism. Reasonable interpolations allowed age composition to be estimated when length data and age–length relationships were measured in different seasons. A growth curve was fitted to the mean length growth using a maximum likelihood method with an assumption of a normal distribution in length variation. Posterior probabilities were constructed with normal distributions according to Bayes’ theorem, and age composition and its confidence limits were reasonably estimated from the posterior probabilities and by bootstrap resampling. The influence of annual fluctuations of population properties was assessed by cross-validation, which was improved by updating the prior probabilities. While the new system was more robust than the age–length key for small numbers of aging data, it was impossible to improve the system by focusing on the length data alone because the correlation between the estimation error and the likelihood calculated from the length data alone was weak.

scholarly journals Evaluation of the effects of otolith sampling strategies and ageing error on estimation of the age composition and growth curve for Pacific bluefin tuna Thunnus orientalis

2019 ◽  
Vol 70 (12) ◽  
pp. 1838 ◽  
Author(s):  
Yi-Jay Chang ◽  
Jhen Hsu ◽  
Jen-Chieh Shiao ◽  
Shui-Kai Chang
Keyword(s):  
Growth Curve ◽  
Sampling Methods ◽  
Growth Parameters ◽  
Estimation Error ◽  
Bluefin Tuna ◽  
Fish Stock ◽  
Pacific Bluefin Tuna ◽  
Thunnus Orientalis ◽  
Age Composition ◽  
Lower Estimation

The age composition of the catch and the growth curve of a stock are fundamentally important in fish stock assessment, but these estimates are subject to various sources of uncertainty. Using the Pacific bluefin tuna (Thunnus orientalis) fisheries in the waters off Taiwan as an example, we developed a Monte Carlo simulation model to evaluate the effects of four otolith sampling methods (random otolith sampling, ROS; fixed otolith sampling, FOS; proportional otolith sampling, POS; and reweighting otolith sampling, REW), and ageing error (bias and imprecision) on estimations of age composition and growth curves. The results indicated that FOS has the lowest sampling accuracy, POS performs the best and that ROS is a more efficient method with lower estimation error. For an imprecise reader, the centre (median) of multiple age reads is a useful method to obtain accurate and precise estimates. Ageing bias had greater effects on the estimation of age composition and growth parameters than ageing imprecision or the selection of otolith sampling methods. In most cases, 500 otoliths should be an adequate sample size and could be the guideline for the biological sampling program of the T. orientalis Catch Documentation Scheme.

An integrated model for growth incorporating tag–recapture, length–frequency, and direct aging data

2004 ◽  
Vol 61 (2) ◽  
pp. 292-306 ◽  
Author(s):  
J Paige Eveson ◽  
Geoff M Laslett ◽  
Tom Polacheck
Keyword(s):  
Growth Curve ◽  
Bluefin Tuna ◽  
Fish Growth ◽  
Data Sources ◽  
Smooth Transition ◽  
Likelihood Method ◽  
Von Bertalanffy ◽  
Length Frequency ◽  
Direct Aging ◽  
Southern Bluefin Tuna

A maximum likelihood method for modelling fish growth is presented that integrates data from three key sources of growth information: tag–recapture studies, length–frequency samples from commercial catches, and direct aging data from hard-parts analyses. Previous studies have almost exclusively modelled growth using only one of these sources of information. Different data sources are often most informative about different portions of the life cycle. The development of an integrated approach allows for the different data sources to complement each other and provide more comprehensive and robust estimates of growth parameters. The integrated method is applied to data sets from southern bluefin tuna (Thunnus maccoyii) using the von Bertalanffy growth curve as well as a more sophisticated growth curve that makes a smooth transition between two von Bertalanffy curves with different growth rate parameters. The latter is found to provide a significantly better fit and supports previous findings that southern bluefin tuna experience a transition in growth during the juvenile stage of life. Many species exhibit a seasonal growth pattern, including southern bluefin tuna for which growth is fastest during the austral summer. A method for incorporating an annual seasonal component into the analysis is described and applied.

scholarly journals A Bayesian hierarchical approach to estimate growth parameters from length data of narrow spread

2019 ◽  
Vol 77 (2) ◽  
pp. 613-623
Author(s):  
Shijie Zhou ◽  
Sarah Martin ◽  
Dan Fu ◽  
Rishi Sharma
Keyword(s):  
Growth Parameters ◽  
Fish Growth ◽  
Hierarchical Approach ◽  
Length Frequency ◽  
Age Classes ◽  
Bayesian Hierarchical ◽  
Gear Selectivity ◽  
Length Data ◽  
Longtail Tuna ◽  
Length Frequency Data

Abstract Estimating fish growth from length frequency data is challenging. There is often a lack of clearly separated modes and modal progression in the length samples due to a combination of factors, including gear selectivity, slowing growth with increasing age, and spatial segregation of different year classes. In this study, we present an innovative Bayesian hierarchical model (BHM) that enables growth to be estimated where there are few distinguishable length modes in the samples. We analyse and identify the modes in multiple length frequency strata using a multinormal mixture model and then integrate the modes and associated variances into the BHM to estimate von Bertalanffy growth parameters. The hierarchical approach allows the parameters to be estimated at regional levels, where they are assumed to represent subpopulations, as well as at species level for the whole stock. We carry out simulations to validate the method and then demonstrate its application to Indian Ocean longtail tuna (Thunnus tonggol). The results show that the estimates are generally consistent with the range of estimates reported in the literature, but with less uncertainty. The BHM can be useful for deriving growth parameters for other species even if the length data contain few age classes and do not exhibit modal progression.

A Procedure for Evaluating the Food Biomass Consumed by a Fish Population

1981 ◽  
Vol 38 (10) ◽  
pp. 1199-1208 ◽  
Author(s):  
Jacek Majkowski ◽  
Kenneth G. Waiwood
Keyword(s):  
Sensitivity Analysis ◽  
Food Consumption ◽  
Atlantic Cod ◽  
Fish Population ◽  
Fish Growth ◽  
Growth Theory ◽  
Food Level ◽  
Fish Food ◽  
Growth Physiology ◽  
Input Parameters

An application of the Ursin fish growth theory for evaluating the food biomass consumed by a fish population is proposed. In this procedure the total catabolism of a fish is treated as dependent on the level of available food. This is the major advantage of the procedure in comparison with that based on the most frequently used Winberg fish growth theory, in which the total catabolism rate is assumed to be independent of the available food level. The Atlantic cod (Gadus morhua) population inhabiting the southern Gulf of St. Lawrence (NAFO subarea 4T) in 1978 is considered in order to illustrate the procedure. The food biomass consumed by the cod population in this area and period is evaluated to be 0.727 million tonnes (the food consumption of the 0th cod age-group is not included). The uncertainty in the estimate of this food biomass due to uncertainties in the input parameters for the procedure is estimated to be in the range of 20% (standard error). Possible ways of improving the results of the procedure are discussed in the light of uncertainties in the input parameters and the sensitivity of the model upon them. It is concluded on the basis of a sensitivity analysis performed for the numerical example of procedure and the foraging theory for fish that the assumption introduced by Winberg can, in certain cases, introduce considerable bias to the procedure results.Key words: fish, food consumption, growth, physiology, bioenergetics, cod, model, sensitivity analysis

Estimation of von Bertalanffy Growth Curve Parameters Using both Length Increment and Age–Length Data

1983 ◽  
Vol 40 (9) ◽  
pp. 1405-1411 ◽  
Author(s):  
G. P. Kirkwood
Keyword(s):  
Growth Curve ◽  
Full Range ◽  
Likelihood Estimation ◽  
Data Sets ◽  
Ratio Test ◽  
Von Bertalanffy ◽  
Predictive Regression ◽  
Length Increment ◽  
Southern Bluefin Tuna ◽  
Length Data

Many fish species cannot be aged directly over their full range of lengths. Therefore, to estimate a growth curve, one often uses length increment data from a mark–recapture experiment, supplemented by whatever age–length data are available. I describe a new method for maximum likelihood estimation of the three von Bertalanffy growth curve parameters, using the length increment and age–length data jointly. Also, I describe a likelihood ratio test for determining whether the same growth curve fits both data sets adequately. The von Bertalanffy growth curve can be taken as a predictive regression with either length or age as the dependent variable. Here, age is taken as the dependent variable, as would be appropriate for estimation of age from length, but only minor modifications are necessary for the more common alternative predictive regression of length on age. As an illustration, the techniques are applied to data for southern bluefin tuna, Thunnus maccoyii.

Marine Growth of Atlantic Salmon (Salmo salar) in the Northwest Atlantic

1972 ◽  
Vol 29 (10) ◽  
pp. 1373-1380 ◽  
Author(s):  
K. Radway Allen ◽  
Richard L. Saunders ◽  
Paul F. Elson
Keyword(s):  
Growth Rate ◽  
Atlantic Salmon ◽  
Salmo Salar ◽  
Growth Curve ◽  
Commercial Fisheries ◽  
The West ◽  
Northwest Atlantic ◽  
Atlantic Salmon Salmo Salar ◽  
Length Data ◽  
Preceding Winter

The fishery for Atlantic salmon in the west Greenland area has provided useful data for the study of marine growth of salmon. Length data from seaward migrating smolts, post-smolts, and 1-, 2-, and 3-sea-winter feeders taken at sea and 1-, 2-, and 3-sea-winter spawners taken by commercial fisheries, angling, and research traps, have been used to construct a tentative growth curve. Fish which spawned after 1 sea-winter (grilse) were smaller at the time of spawning than fish of the same smolt-class which had not matured but were still actively feeding at sea. Similarly, 2-sea-winter spawners were smaller than salmon which would not have returned until after a third sea winter. The growth rate of salmon during the summer prior to spawning was lower than that of fish which would have spawned a year or more later and appears to be about the same as that during the preceding winter.

Boosting Method for Local Learning in Statistical Pattern Recognition

2008 ◽  
Vol 20 (11) ◽  
pp. 2792-2838 ◽  
Author(s):  
Masanori Kawakita ◽  
Shinto Eguchi
Keyword(s):  
Estimation Error ◽  
Approximation Error ◽  
Real Data ◽  
Classification Performance ◽  
Classification Rule ◽  
Likelihood Method ◽  
Data Sets ◽  
Bayes Risk ◽  
Classification Problems ◽  
Boosting Method

We propose a local boosting method in classification problems borrowing from an idea of the local likelihood method. Our proposal, local boosting, includes a simple device for localization for computational feasibility. We proved the Bayes risk consistency of the local boosting in the framework of Probably approximately correct learning. Inspection of the proof provides a useful viewpoint for comparing ordinary boosting and local boosting with respect to the estimation error and the approximation error. Both boosting methods have the Bayes risk consistency if their approximation errors decrease to zero. Compared to ordinary boosting, local boosting may perform better by controlling the trade-off between the estimation error and the approximation error. Ordinary boosting with complicated base classifiers or other strong classification methods, including kernel machines, may have classification performance comparable to local boosting with simple base classifiers, for example, decision stumps. Local boosting, however, has an advantage with respect to interpretability. Local boosting with simple base classifiers offers a simple way to specify which features are informative and how their values contribute to a classification rule even though locally. Several numerical studies on real data sets confirm these advantages of local boosting.

V.—A Fish Cultivation Experiment in an Arm of a Sea-loch. V. Fish Growth in Kyle Scotnish

1950 ◽  
Vol 64 (1) ◽  
pp. 109-135 ◽  
Author(s):  
Fabius Gross
Keyword(s):  
Fish Population ◽  
Fish Growth ◽  
Native Fish ◽  
Otter Trawl ◽  
Large Numbers ◽  
Cultivation Experiment

Fishing with both a shore seine net and a small otter trawl in the summer of 1943 and the spring of 1944 showed that the native fish population was very sparse. No estimate of the size of fish population was attempted in 1944. Estimates undertaken in 1945 and 1946 will be discussed later (p. 120). Fishing in the period June to September 1945 produced small numbers of fish tabulated on p. 110, and some plaice and flounders which will be dealt with separately.In addition there were fairly large numbers of gobies, sticklebacks, pipe fish and sand eels which, however, were not usually recorded in our catches. Shoals of young herring appeared only in the summer and autumn of 1945.

Are Growth Parameters Estimated from Tagging and Age–Length Data Comparable?

1988 ◽  
Vol 45 (6) ◽  
pp. 936-942 ◽  
Author(s):  
R. I. C. C. Francis
Keyword(s):  
Growth Rate ◽  
Growth Rates ◽  
Growth Parameters ◽  
Growth Models ◽  
Maximum Length ◽  
Fish Growth ◽  
Data Sets ◽  
Von Bertalanffy ◽  
Length Data ◽  
The Mean

The two most common ways of estimating fish growth use age–length data and tagging data. It is shown that growth parameters estimated from these two types of data have different meanings and thus are not directly comparable. In particular, the von Bertalanffy parameter l∞ means asymptotic mean length at age for age–length data, and maximum length for tagging data, when estimated by conventional methods. New parameterizations are given for the von Bertalanffy equation which avoid this ambiguity and better represent the growth information in the two types of data. The comparison between growth estimates from these data sets is shown to be equivalent to comparing the mean growth rate of fish of a given age with that of fish of length equal to the mean length at that age. How much these growth rates may differ in real populations remains unresolved: estimates for two species of fish produced markedly different results, neither of which could be reproduced using growth models. Existing growth models are shown to be inadequate to answer this question.

Sign in / Sign up

Export Citation Format

Share Document