TropFishR

Performance evaluation of data-limited, length-based methods is instrumental in determining and quantifying their accuracy under various scenarios and in providing guidance about model applicability and limitations. We conducted a simulation-estimation analysis to compare the performance of four length-based stock assessment methods: length-based Thompson and Bell (TB), length-based spawning potential ratio (LBSPR), length-based integrated mixed effects (LIME), and length-based risk analysis (LBRA), under varying life history, exploitation status, and recruitment error scenarios. Across all scenarios, TB and LBSPR were the most consistent and accurate assessment methods. LBRA is highly biased, but precautionary, and LIME is more suitable for assessments with time-series longer than a year. All methods have difficulties when assessing short-lived species. The methods are less accurate in estimating the degree of recruitment overfishing when the stocks are severely overexploited, and inconsistent in determining growth overfishing when the stocks are underexploited. Increased recruitment error reduces precision but can decrease bias in estimations. This study highlights the importance of quantifying the accuracy of stock assessment methods and testing methods under different scenarios to determine their strengths and weaknesses and provides guidance on which methods to employ in various situations.

The determination of rates of body growth is the first step in many aquatic population studies and fisheries stock assessments. ELEFAN (Electronic LEngth Frequency ANalysis) is a widely used method to fit a growth curve to length-frequency distribution (LFD) data. However, up to now, it was not possible to assess its accuracy or the uncertainty inherent of this method, or to obtain confidence intervals for growth parameters within an unconstrained search space. In this study, experiments were conducted to assess the precision and accuracy of bootstrapped and single-fit ELEFAN-based curve fitting methods, using synthetic LFDs with known input parameters and a real data set of Abra alba shell lengths. The comparison of several types of bootstrap experiments and their outputs (95% confidence intervals and confidence contour plots) provided a first glimpse into the accuracy of modern ELEFAN-based fit methods. The main components of uncertainty (precision and reproducibility of fit algorithms, seed effects, sample size and matrix information content) could be assessed from partial bootstraps. Uncertainty was mainly determined by LFD matrix size (months x size bins), total number of non-zero bins and the sampling of large-sized individuals. A new pseudo-R² index for the goodness-of-fit of von Bertalanffy growth models to LFD data is proposed. For a large, perfect synthetic data set, pseudo-R²Phi’ was very high (88 to 100%), indicating an excellent fit of the growth model. The small Abra alba data set showed a low pseudo-R²Phi’, from to 54% to 68%, indicating the need for more samples (length measurements) and a larger LFD data matrix. New, robust, bootstrap-based methods for curve fitting are presented and discussed. This study demonstrates a promising new path for length-based analyses of growth and mortality in natural populations, which are the basis for a suite of methods that are included in the new fishboot package.

Small-scale fisheries (SSF) contribute to approximately half of the total landings of tuna and tuna-like species in the Indian Ocean and are an important form of employment and source of protein. Research into the properties and dynamics of SSF in East Africa are important for the assessment and sustainable management of fish stocks,however, detailed fisheries data are often inadequate or absent. Fisheries-dependent data on driftnet fisheries in Zanzibar, Tanzania, was collected during the northeast monsoon seasons in 2014 and 2015. The data describes the properties of the driftnet fisheries and allows for comparisons of the length composition of the landings of the SSF with large-scale industrial fisheries (IF) fishing in Tanzania’s Exclusive Economic Zone (EEZ). This data also facilitates the calculation of stock indicators for the five most abundant tuna and tuna-like species landed in Zanzibar.Results show that the two fisheries (SSF and IF) exploit the same stocks, and landings are representative of a similar length composition, while operating in different parts of Tanzania’s EEZ. High exploitation rates, above reference levels for all species were calculated, in agreement with official assessments by the IOTC, and suggest that calls for the expansion of the SSF should be reconsidered. The assessment and management of straddling stocks are dis-cussed, as well as solutions to challenges faced by local observer programmes.

Small-scale multi-gear fisheries contribute half of global fisheries landings but are generally data-poor, hindering their assessment and management. Aiming to overcome various existing challenges, we used two complementary length-based approaches to assess the status of three main target species in the small-scale fisheries of Eastern Pacific countries: Spotted rose snapper Lutjanus guttatus, Pacific sierra Scomberomorus sierra, and Pacific bearded Brotula clarkae, using length-frequency catch data (LFCD) from the Colombian Pacific coast. Two data sources – official governmental data and community-based monitoring from a non-government organization – were used to estimate two sets of stock indicators: one based on the derivation of growth and mortality parameters from modal progression, catch curve analysis and a yield-per-recruit model using TropFishR; and the second based on the relative contribution of fish sizes with regard to proposed reference values for healthy stocks. Growth estimates differed between data sources and exhibited large confidence intervals, indicating an overall high uncertainty underlying the LFCD revealed through a novel bootstrapped approach. Estimated values of stock indicators, exploitation rate, fishing mortality and size-proportions converged in suggesting a state of heavy to over-exploitation for the three assessed species, although differences were observed among data sources that we attribute mainly to fisheries selectivity and sampling design. In order to improve future assessments of stocks in multi-gear and data-poor contexts, estimations of fleet-specific selectivity should be used to reconstruct LFCD prior to analyses. Additionally, sampling design should be based on fishing effort distribution among gears and areas and, when feasible, fishery-independent data on stock conditions should be included.

The determination of rates of body growth is the first step in many aquatic population studies and fisheries stock assessments. ELEFAN (Electronic LEngth Frequency ANalysis) is a widely used method to fit a growth curve to length-frequency distribution (LFD) data. However, up to now, it was not possible to assess its accuracy or the uncertainty inherent of this method, or to obtain confidence intervals for growth parameters within an unconstrained search space. In this study, experiments were conducted to assess the precision and accuracy of bootstrapped and single-fit ELEFAN-based curve fitting methods, using synthetic LFDs with known input parameters and a real data set of Abra alba shell lengths. The comparison of several types of bootstrap experiments and their outputs (95% confidence intervals and confidence contour plots) provided a first glimpse into the accuracy of modern ELEFAN-based fit methods. The main components of uncertainty (precision and reproducibility of fit algorithms, seed effects, sample size and matrix information content) could be assessed from partial bootstraps. Uncertainty was mainly determined by LFD matrix size, total number of non-zero bins and the sampling of large-sized individuals. A new pseudo-Rsquared index for the goodness-of-fit of VBGF models to LFD data is proposed. For a large, perfect synthetic data set, pseudo-RsquaredPhi was very high (88 to 100%), indicating an excellent fit of the VBGF model. The small Abra alba data set showed a low pseudo-RsquaredPhi, from to 54% to 68%, indicating the need for more samples and a larger LFD data matrix. New, robust, bootstrap-based methods for curve fitting are presented and discussed. This study demonstrates a promising new path for length-based analyses of growth and mortality in natural populations, which are the basis for a new suite of methods that are included in the new fishboot package.

Electronic length frequency analysis (ELEFAN) is a system of stock assessment methods using length-frequency (LFQ) data. One step is the estimation of growth from the progression of LFQ modes through time using the von Bertalanffy growth function (VBGF). The option to fit a seasonally oscillating VBGF (soVBGF) requires a more intensive search due to two additional parameters. This work describes the implementation of two optimisation approaches (“simulated annealing” and “genetic algorithm”) for growth function fitting using the open-source software “R.” Using a generated LFQ data set with known values, the accuracy of the soVBGF parameter estimation was evaluated. The results indicate that both optimisation approaches are capable of finding high scoring solutions, yet settings regarding the initial restructuring process for LFQ bin scoring (i.e. “moving average,”) and the fixing of the asymptotic length parameter (L∞) are found to have significant effects on parameter estimation error. An outlook provides context as to the significance of the R-based implementation for further testing and development, as well as the general relevance of the method for data-limited stock assessment.

The R package TropFishR is a new analysis toolbox compiling single-species stock assessment methods specifically designed for data-limited fisheries analysis using length-frequency data. It includes methods for (i) estimating biological stock characteristics such as growth and mortality parameters, (ii) exploring technical aspects of the fisheries (e.g. exploitation rate and selectivity characteristics), (iii) assessing size and composition of a fish stock by means of virtual population analysis (VPA), and (iv) assessing stock status with yield prediction and production models. This paper introduces the package and demonstrates the functionality of a selection of its core methods. TropFishR modernises traditional stock assessment methods by easing application and development and by combining it with advanced statistical approaches. This article is protected by copyright. All rights reserved.

Fish stock assessment methods and fisheries models based on the FAO Manual ”Introduction to tropical fish stock assessment” by P. Sparre and S.C. Venema <http://www.fao.org/docrep/W5449E/W5449E00.htm>. Focus is the analysis of length-frequency data and data poor fisheries.