Quantitative Fisheries Research Articles

DESIGNING FISHERY MODELS: A PERSONAL ADVENTURE

ABSTRACT. This paper describes my experiences designing fishery models, starting from a mathematical background in the differential equations of theoretical physics. Three examples from my early research, cited by Quinn in the lead article for this issue, illustrate a historical approach to model design. Although such analytical results provide useful tools for thought, they sometimes gloss over important assumptions and limitations. I describe the series of questions that led me from simple models to a more complete statistical framework, involving state space models and Bayes statistics. Modern fishery models often grow into complex structures that depend on numerous arbitrary choices about underlying deterministic processes, process error, and measurement error. Given this inherent ambiguity and uncertainty, I discuss scientific limits to quantitative fishery models and future prospects for devising robust management algorithms.

Extraction and comparison of acoustic backscatter from a calibrated multi- and single-beam sonar

Abstract Multi-beam sonar is potentially a powerful analytical tool for investigating the acoustic properties and behaviour of fish in relation to quantitative fisheries research. The SIMRAD SM2000 is a 200 kHz multi-beam sonar employing an 80-element array to transmit and synthesize, electronically, 128 receive beams (20°×2.2°) over a 180° arc simultaneously. Once calibrated, such systems enable the extraction of acoustic target strength and volume backscattering from an extended 3D ocean volume. We present an overview of the theoretical framework for the calibration of a multi-beam sonar, and then compare the acoustic backscatter from a calibrated single-beam 50 kHz echosounder with selected beams from a sphere-calibrated multi-beam sonar. Both systems recorded acoustic data from Atlantic herring contained within a weir, as the fish passed beneath the transducers. Specifically, we examine the relationship between the area-backscattering strength (Sa) from the single-beam system with the nadir beam (beam 63) of the SM2000 sonar. In addition, data are presented on the observed variability in Sa with target aspect for off-vertical angles from 15° to 60° in 15° intervals. Non-standard synthesized SM2000 beam widths are explored for both calibration and field datasets. The implications for biomass estimation are also discussed.

Estimating Parameters of Derived Random Variables: Comparison of the Delta and Parametric Bootstrap Methods

In quantitative fisheries research, analysts often need to estimate the parameters of derived random variables that cannot be observed directly and are instead computed from other observed random variables. The delta method, which is based on the Taylor series, has been widely used for approximating the mean and variance of derived variables. This paper applies a computer simulation method that involves the parametric bootstrap method for estimating the mean and variance of derived random variables. The results from these two methods are compared with the true values (when they can be computed) or with each other. Several types of commonly used functions in quantitative fisheries studies are examined, namely, the exponential, logarithmic, product, quotient, and asymptotic functions. The parametric bootstrap method leads to fairly accurate estimates for the mean and variance of the derived variables. The delta method provides reasonably good estimates when the variation (expressed as the coefficient of variation, CV, which is defined as 100·SD/mean) of the underlying variable is small; however, the deviation from the true values or the bootstrap results increases as the variations of the underlying observed variables increase. The impact is more severe for the variance estimator than for the mean estimator. The delta method under- or overestimates the variance by about 10% when the CVs of the underlying variables approach the following values: 25% for the function exp(x), 20% for loge(x), 50% for xy, 15% for x/y and 1/x, 25% for loge(x/y), and 30% for x(1—exp(-yt)). It is recommended that the bootstrap method be used for estimating the variance of a derived random variable when the observed variables have a large variance.

Three-dimensional visualization of acoustic backscattering and models by alewife

The advent of nontraditional underwater acoustic instrumentation (e.g., multibeam and sector-scanning sonar) and incorporating fish behavior in quantitative fisheries assessment require that backscatter be measured and modeled at a variety of aspect angles. Acoustic backscattering measurements were obtained in a large laboratory tank using live alewife (Alosa pseudoharengus). An individual alewife for each series of measurements was tethered and rotated in two planes of orientation (dorsal/ventral and lateral). The alewife were insonified with a broadband (40–100 kHz) chirp signal and bistatic scattering geometry was used. Backscattering amplitudes for all angles of orientation (3-D scattering ambit) were modeled using a Kirchhoff ray-mode model [C. S. Clay and J. K. Horne, J. Acoust. Soc. Am. 96, 1661–1668 (1994)] and digital images of the fish body and swimbladder morphometry. Visualization of the scattering ambit provides a quantitative examination of the effects of fish orientation on echo amplitude. Comparisons of backscattering amplitudes from the model and measurements along the dorsal/ventral and lateral planes are given. The utility of broadband measurements for fish backscattering amplitude measurements and the integration of acoustic models in fisheries assessments are discussed. [Work supported by ONR, NOAA/NMFS, and CICOR.]

Evidence for biotic control of Australian reservoir phytoplankton communities and the potential for applied biomanipulation

AbstractAlthough some concerns have been expressed with respect to the applicability of biomanipulation to Australian reservoirs and other inland waters, no major obstacles were found in this review. The previous doubts stimulated research, which has provided evidence for unexpected mechanisms of food web interactions, such as positive effects of small copepods on planktonic algae, higher than usual grazing potential of cladoceran communities, and direct positive effects of planktivores on phytoplankton. Further research into zooplankton‐cyanobacteria interactions should indicate how widespread is the previously reported ability of some southern hemisphere crustaceans to benefit at a population level from feeding on cyanobacteria. Quantitative fisheries acoustics, which has not been widely used in biomanipulation trials before, is a promising new tool in the assessment of the biomass of planktivorous fish in the pelagic zone of reservoirs.

Fisheries harvest and standing stock in a Hawaiian Bay

An 18-month study at Hanalei Bay, Kauai provided unusually valuable quantitative fishery data and concurrent estimates of natural density of reef fish for a complete, small, subsistence/recreational/artisanal fishery on a small, remote island in the tropical mid-Pacific. Catch and effort data were collected using a stratified random sampling design. The commercial surround net fishery targeted mainly coastal pelagic species and accounted for over 70% of the catch. Most other fishing was for recreation or subsistence. Line fishing from shore and boats accounted for much of the effort but produced some of the lowest catches per unit effort (CPUE). The cast net and spear fisheries had CPUE values of more than 1 kg per gear-hour and caught a wide diversity of reef-associated species. Standing stock of reef-associated fishes was estimated from 516 underwater visual censuses. Mean biomass ranged from over 16 kg/100 m 2 in shallow complex habitats to less than 1 kg/100 m 2 in monotypic reef flat habitats. These densities, which are higher than those reported from some more populated, similar areas in Hawaii, may reflect lower fishing pressure in Hanalei. Yield estimates from Hanalei Bay are very low compared to those from many other locations in the Pacific. The small fraction of the overall standing stock that is harvested annually (~ 1.3%) suggests that most of the fish community is not being severely overfished. Parrotfishes, goatfishes, and surgeonfishes were important components of the fisheries and of the censused communities, suggesting that these target species have not been seriously depleted. The small sizes at which some valued species are caught is a matter of concern for management of these stocks.

Behavior of Fisheries Quantitative Fisheries Stock Assessment: Choice Dynamics and Uncertainty Ray Hilborn Carl J. Walters

Behavior of Fisheries Quantitative Fisheries Stock Assessment: Choice Dynamics and Uncertainty Ray Hilborn Carl J. Walters

Quantitative Fisheries Stock Assessment: Choice, Dynamics and Uncertainty.Ray Hilborn , Carl J. Walters

<i>Quantitative Fisheries Stock Assessment: Choice, Dynamics and Uncertainty.</i>Ray Hilborn , Carl J. Walters

Quantitative fisheries stock assessment: Choice, dynamics and uncertainty

Quantitative fisheries stock assessment: Choice, dynamics and uncertainty

Models for fisheries management

Realization of the massive potential harvests from the oceans, coupled with disastrous collapses of substantial fisheries has seen the development over the last 25 years of a quantitative fisheries management science. As an aid to management, models are used to represent fish populations and the effects of exploitation on them. Basic models currently used for fisheries management are reviewed. The models discussed range from the simple surplus production model through to the analytical model of Beverton and Holt incorporating age structure and finally to Kicker's stock-recruitment model. The models are used to examine the dynamics and stability of fish stock when subjected to different exploitation policies. Some of the problems for management in interfacing real fisheries data with the mathematical models are discussed. The goals of management, including catch, fishing effort and economics (e.g. price, cost and inflation) are quantified to produce an objective function which may be used to determine optimal strategies. The optimal harvesting strategies are examined for each model, with a view to developing a general workable robust harvesting policy, independent of the model used.

Detection and quantification of fish populations by ultrasonic echo ranging

The discovery that fish schools could be detected by ultrasonic echo ranging led to early utilization of hydroacoustic equipment by fisherman and fisheries scientists. However, the full potential of ultrasonic echo ranging for quantitative fisheries investigations has only recently begun to be realized. At the University of Washington a concerted effort to develop and apply hydroacoustic methods in the study of fish populations was initiated in the late 1960's. Investigations have been conducted on the relationship between echo intensity and fish abundance and on scattering characteristics of fish targets. Equipment and techniques which were developed as part of this research are now being applied by fisheries managers on several fish stocks in the northeastern Pacific and in scientific investigations of the role of fish in both freshwater and marine ecosystems. [Work supported by Sea Grant and IDOE.]