Canadian Journal of Fisheries and Aquatic Sciences

Published by NRC Research Press
Online ISSN: 1205-7533
Conference Paper
The usefulness of side-looking sonar in fisheries assessment work can be greatly increased by employing array phase information to estimate the echo arrival direction. Vertical and horizontal coordinates (perpendicular to the beam direction) can be measured, permitting beam pattern corrections, tracking from ping to ping, and depth determination. Acoustic data have been gathered during cruises in the Northeast Pacific, Caribbean, and Northeast Atlantic using both side- and down-looking transducers. The down-looking transducer was mounted in a towed body and interfaced with a simrad EK 120 scientific sounder. The side-looking transducer was part of a towed measurement system that provided phase-coherent recording of signals received by transducer quadrants.[1] This system contained sensors for speed, depth, pitch, and roll. Depth distributions have been computed using both side and down-looking data. Estimates of nekton biomass versus vertical and horizontal distance were made by echo integration of down-looking data. Target strength distributions have been computed from the side-looking data, with and without beam loss corrections, which produce shifts in the distributions on the order of 3 dB. Echo-to-Echo fluctuations in side-looking depth measurements correspond to a standard deviation of 10 -20 m for the Northeast Atlantic schools, which are assumed to be Blue Whiting. Information is also presented on Northeast Pacific rockfish schools and on unknown species in the Western Caribbean Sea.
The failures of traditional target-species management have led many to propose an ecosystem approach to fisheries to promote sustainability. The ecosystem approach is necessary, especially to account for fishery-ecosystem interactions, but by itself is not sufficient to address two important factors contributing to unsustainable fisheries —inappropriate incentives bearing on fishers, and the ineffective governance that frequentlyexists in commercial, developed fisheries managed primarily by total harvest limits and input-controls. We contend that much greater emphasis must be placed on fisher motivation when managing fisheries. Using evidence from more than a dozen ‘natural experiments’ in commercial fisheries, we argue that incentive-based approaches that better specify community, individual harvest, or territorial rights and also price ecosystem services — coupled with public research, monitoring and effective oversight — promote sustainable fisheries.
Fisheries managers in the United States are required to identify and mitigate the adverse impacts of fishing activity on essential fish habitat (EFH). There are additional concerns that the viability of noncommercial species, animals that are habitat dependent and/or are themselves constituents of fishery habitat may still be threatened. We consider a cap-and-trade system for habitat conservation, individual habitat quotas for fisheries, to achieve habitat conservation and species protection goals cost effectively. Individual quotas of habitat impact units (HIUs) would be distributed to fishers with an aggregate quota set to maintain a target habitat “stock” of EFH conservation. Using a dynamic, spatially explicit fishery simulation model we explore the efficiency and cost effectiveness of an IHQ policy versus alternative marine protected area (MPA) configurations, at reducing the risk of extinguishing a habitat dependent species of unknown spatial distribution. Our findings indicate that an IHQ policy with a conservatively established habitat target is better suited to the protection of non-target species than a rotating or fixed MPA policy.
Percentage of overall salmon catch from each area, 1984-1993 (DFO 1993).
Relative catch levels for the different areas, 1984-1993 (DFO 1993).
Canada's Pacific salmon fisheries are again in a crisis of declining stocks, excess fishing effort and poor returns. Changes in fisheries management which may help resolve some of these problems include licence buy-backs, individual transferable quotas, additional effort controls, single gear licensing and area licensing or combination thereof. This paper proposes an alternative fishers' individual salmon harvesting rights which provide fishers with rights to bid for the privilege of participating in salmon fishing openings.
Total and relative efforts and achievable abundances in managed fisheries. ( a ) With a single métier in a fishery, stocks’ relative equilibrium abundances are constrained (the vulnerability constraint (red line)) by their relative vulnerabilities to the métier. ( b ) As a result, with few métiers in a multispecies fishery it is often impossible to simultaneously achieve target abundances for multiple stocks, instead trading off overexploiting some with under-exploiting others. ( c ) With multiple métiers, management influencing both their relative and total effort levels can produce any set of equilibrium stock abundances within the region (shaded) bounded by the vulnerability constraints each would produce in isolation (red and blue lines). 
Fleet diversity increases achievable yields and profits in managed fisheries. Results of a five-stock stochastic simulation of average relationships between fleet diversity and achievable yields and profits with (a) optimal management for obtaining maximum achievable yield (MAY) and (b) optimal management for obtaining maximum achievable profit (MAP). Mean (a) MAYs and (b) MAPs are shown (black) along with the mean number of extinctions required for achieving MAY or MAP (grey). Each point represents a sample of 500 models with randomly chosen parameter values. Vertical lines indicate standard errors. Points in panel (a) showing simulation results with diverse métiers only (black squares, black dashed line) and those showing results with diversity in both métiers and efficiency (black circles, black solid line) are nearly identical. This is due to the fact that diversity in efficiency among fishing units has no effect on achievable yields.
Consequences of fleet diversification in open-access fisheries. ( a ) A stochastic simulation showing the frequency distribution of equilibria ( N *) in a 2 stock fishery in which 1, 2, and 3 fishing units respectively (from left to right) were randomly drawn 15 000 times from a uniform distribution of all possible métiers ( m ij ~ U[0,1]) and 
Figure A1e-predator and essential prey
Biological diversity is known to play an important role in generating and maintaining ecosystem productivity and other functions, and has consequently become a central focus of many efforts to preserve ecosystem services. Theoretical parallels suggest the diversity of fishing fleets may have a similarly important role in determining the productivity and ecological impacts of fisheries, but this possibility has rarely been explored. Here I present theoretical analyses showing that the diversity of métiers – combinations of technology, target species, and fishing grounds – and technical efficiencies in a fishing fleet have important impacts on the productivity, profitability, and ecological impacts of fisheries, particularly mixed-stock or multispecies fisheries. Diversification of métiers can increase yields and reduce threats to weak stocks in both managed and unmanaged multispecies fisheries. Diversification of technical efficiencies creates opportunities for larger profits in managed fisheries, but often decreases yields and worsens impacts on weak stocks in unmanaged fisheries. These results suggest that the potential impact of management may be highest in fisheries with diverse fleets.
Innovation policy is increasingly informed from the perspective of a national innovation system (NIS), but, despite the fact that research findings emphasize the importance of national differences in the framing conditions for innovation, policy prescriptions tend to be uniform. Justifications for innovation policy by organizations such as the OECD generally relate to notions of market failure, and the USA, with its focus on the commercialization of public sector research and entrepreneurship, is commonly portrayed as the best model for international emulation. In this paper we develop a broad framework for NIS analysis, involving free market, coordination and complex-evolutionary system approaches. We argue that empirical evidence supporting the hypothesis that the ‘free market’ can be relied upon to promote innovation is limited, even in the USA, and the global financial crisis provides us with new opportunities to consider alternatives. The case of Australia is particularly interesting: a successful economy, but one that faces continuing productivity and innovation challenges. Drawing on information and analysis collected for a major review of Australia’s NIS, and the government’s 10-year plan in response to it, we show how the free market trajectory of policy-making of past decades is being extended, complemented and refocused by new approaches to coordination and complex-evolutionary system thinking. These approaches are shown to emphasize the importance of systemic connectivity, evolving institutions and organizational capabilities. Nonetheless, despite the fact that there has been much progress in this direction in the Australian debate, the predominant logic behind policy choices still remains one of addressing market failure, and the primary focus of policy attention continues to be science and research rather than demand-led approaches. We discuss how the development and elaboration of notions of systems failure, rather than just market failure, ca
This paper presents a Bayesian approach to fisheries stock assessment using the delay difference model to describe nonlinear population dynamics. Given a time series of annual catch and effort data, models in the Deriso-Schnute family predict exploitable biomass in the following year from biomass in the current and previous year and from past spawning stock. A state-space model is used as it allows to incorporate both random errors in the biomass dynamics equations and the observations. Because the biomass dynamics are nonlinear, the common Kalman filter is generally not applicable for parameter estimation. However, it will be demonstrated that the Bayesian approach can handle any form of nonlinear relationship in the state and observation equations as well as realistic distributional assumptions. Difficulties with posterior calculations are overcome by the Gibbs sampler in conjunction with the adaptive rejection Metropolis sampling algorithm. Introduction Delay difference models bri...
This paper illustrates the ease with which Bayesian nonlinear state-space models can now be used for practical fisheries stock assessment. Sampling from the joint posterior density is accomplished using Gibbs sampling via BUGS, a freely available software package. By taking advantage of the model representation as a directed acyclic graph, BUGS automates the hitherto tedious calculation of the full conditional posterior distributions. Moreover, the output from BUGS can be read directly into the software CODA for convergence diagnostics and statistical summary. We illustrate the BUGS implementation of a nonlinear non-normal state-space model using a Schaefer surplus production model as a basic example. This approach extends to other assessment methodologies, including delaydifference and age-structured models. Introduction State-space models are among the most powerful tools for dynamic modeling and forecasting (Fahrmeir and Tutz 1994). They have started to enjoy an increasing popular...
Variation among individuals in specific growth rate (SGR), feeding, and two measures of swimming performance and their possible interrelationships were investigated in juvenile rainbow trout (Oncorhynchus mykiss) kept in groups on either satiation or half-satiation rations. Maximum sustainable velocity (U crit ) was measured as an index of aerobic swimming performance and stamina (fatigue time in a fixed-velocity test at 6 body lengthss --1 ) as an index of anaerobic performance. Individual performance in both of these tests was found to be significantly repeatable. Trout fed on half-satiation ration exhibited significantly lower mean values of SGR and body size and higher levels of aggression-related fin damage, but no significant differences in stamina, relative or absolute U crit , glycogen content, or plasma cortisol. However, in these fish, there was a significant negative relationship between SGR and relative U crit , a significant positive relationship between SGR and stamina, and a significant positive relationship between SGR and total daily meal. None of these relationships were seen in fish fed to satiation. Plasma cortisol and tissue glycogen stores were not related to SGR. These results indicate that under the intensified competition of restricted ration, there are trade-offs between growth, feeding, and different types of swimming performance. Rsum : La variation entre individus du taux de croissance spcifique (TCS), de l'alimentation et de deux mesures de la performance natatoire, de mme que les relations possibles entre ces paramtres, ont t tudies chez des truites arc-en-ciel (Oncorhynchus mykiss) juvniles rparties en deux groupes : un premier groupe nourri satit et un deuxime groupe moiti rassasi. La vitesse soutenable maximale (U cr...
Map of insular Newfoundland showing localities where squid were sampled for stomach analysis.
Distribution of length, based on otoliths recovered from squid stomachs, for five fish species (fork length for Atlantic cod, total length for all other species). Ranges of length are given in parentheses. Two specimens of sand lance larger than 210 mm are not represented.
Monthly mean number of otoliths per stomach (for all stomachs examined) by locality and year.
: Fish otoliths were collected from stomachs of short-fined squid (Illex illecebrosus) at 11 coastal Newfoundland localities over 11 years during 1980--1993. Most oltoliths were of young-of-the-year Atlantic cod (Gadus morhua) and they were common at all localities after July. Adult capelin (Mallotus villosus) otoliths were common early in the season at a southern locality whereas otoliths of juvenile sand lance (Ammodytes sp.) were common later in the season at the northern localities. Other fish that were relatively well represented in only a few otolith collections were Arctic cod (Boreogadus saida), Atlantic herring (Clupea harengus), redfish (Sebastes sp.), and unidentified hake (Gadidae). Arctic cod otoliths were unusually prominent in only two years and from northernmost localities. Redfish otoliths were especially prevalent in the only collection from the south coast of Newfoundland. Paramount among the biases inherent in using this approach to estimate the effects of p...
Schematic of a multispecies phosphorus model (modified from Chapra and Reckhow 1983). 
Probability network model for the Neuse River estuary. Probabilities adjacent to the ovals and diamond are hypothetical examples. NR, nitrogen load reduction; MMS, March-May daily streamflow at Kinston; JAS, June-August daily streamflow at Kinston; TN, total nitrogen concentration. 
Conditional probabilities for intermediate variables % N load reduction, TN concentration, and summer streamflow. 
It is a common strategy in surface water quality modeling to attempt to remedy predictive inadequacies by incorporating additional mechanistic detail into the model. This approach reflects the reasonable belief that enhanced scientific understanding of basic processes can be used to improve predictive modeling. However, nature is complex, and even the most detailed simulation model is extremely simple in comparison. At some point, additional detail exceeds our ability to simulate and predict with reasonable error levels. In those situations, an attractive alternative may be to express the complex behavior probabilistically, as in statistical mechanics, for example. This viewpoint is the basis for consideration of Bayesian probability networks for surface water quality assessment and prediction. To begin this examination of Bayes nets, some simple water quality examples are used for the illustration of basic ideas. This is followed by discussion of a set of proposed probability network ...
(a) Reflected light image of a transverse-sectioned saggitae otolith after ion microprobe analysis from a 4+-year-old bluegill (BLG-3), Lepomis macrochirus, collected from Crampton Lake, Wisconsin, in September 2006. The surface is coated with gold thin film, and analysis pits appear as uniform dark spots. δ 18 O pits are along the top of the sampling traverse and are slightly larger (15 µm) than δ 13 C pits (10 µm). Arrows indicate interpretations of annual growth bands. (b) Scanning electron microscope (SEM) image (backscattered electron mode) of δ 18 O and δ 13 C analysis pits at the otolith origin where daily growth increment widths are 1040 µm. (c and d) SEM images from two points within the last year of growth illustrate how daily increment width, an indicator of fish growth rate, influences the number of days incorporated in an analysis.
Otoliths provide information about an individual fish's environment at ecologically relevant time scales. We used ion microprobe analysis to produce high-resolution delta C-13 and delta O-18 time series from two age-4 bluegill (Lepomis macrochirus) otoliths, which provided insight into fish behavior and otolith fractionation processes. Scanning electron microscope images revealed delta C-13 and delta O-18 pit diameters of 10 and 15 mu m, respectively, corresponding to 1-5 and 29 daily increments during rapid otolith growth and 6-9 and 12-25 increments near annual otolith growth checks. Spot-to-spot reproducibility (1 SD) of the calcite standards was <0.2 parts per thousand for delta O-18 and <0.4 parts per thousand for delta C-13 and was small enough to resolve a change in a fish's ambient temperature of approximately 1 degrees C. A whole-take C-13 addition experiment elevated the delta C-13 of the lake's dissolved inorganic carbon for 56 days during the summer of 2005. Mixing model results indicated that the proportion of dietary carbon in otoliths (M) was similar for both fish (BLG-3, M = 0.45; BLG-12, M = 0.35), but the relation between M and proxies of metabolic rate differed between fish. Otolith stable isotope analysis by ion microprobe can reveal the environmental history of an individual fish and contribute to our understanding of processes that influence isotope ratio fractionation in otoliths.
Ordination comparison of the two most similar years, showing correspondence analysis (CA) ordinations of (a) 1990 and (b) 1991, canonical correspondence analysis (CCA) ordinations of (c) 1990 and (d) 1991, and (e) translated CA ordinations of both 1990 and 1991 after Procrustes analysis. Lines in (e) indicate residual differences between rotated ordination plots. Taxa indicated in (a) and (b) are Amphipoda (Amp), Anisoptera (Ani), Bivalvia (Biv), Ceratopogonidae (Cer), Chironomidae (Chi), Coleoptera (Col), Ephemeroptera (Eph), Gastropoda (Gas), Hemiptera (Hem), Lepidoptera (Lep), Nematoda (Nem), Oligochaeta (Oli), Trichoptera (Tri), and Zygoptera (Zyg). Lake codes in (a) – (e) are as indicated in Table 1. DOC, dissolved organic carbon; Alk, alkalinity.  
Ordination comparison of the two most dissimilar years, showing correspondence analysis (CA) ordinations of (a) 1997 and (b) 1998, canonical correspondence analysis (CCA) ordinations of (c) 1997 and (d) 1998, and (e) translated CA ordinations of both 1997 and 1998 after Procrustes analysis. Lines in (e) indicate residual differences between rotated ordination plots. Taxon codes in (a) and (b) are Amphipoda (Amp), Anisoptera (Ani), Bivalvia (Biv), Ceratopogonidae (Cer), Chironomidae (Chi), Coleoptera (Col), Ephemeroptera (Eph), Gastropoda (Gas), Hemiptera (Hem), Lepidoptera (Lep), Nematoda (Nem), Oligochaeta (Oli), Trichoptera (Tri), Zygoptera (Zyg), Decapoda (Dec), Hirudinea (Hir), and Trombidiformes (Tro). Lake codes in (a) – (e) are as indicated in Table 1. DOC, dissolved organic carbon; Alk, alkalinity.  
Few studies of biological recovery from acidification have dealt with community responses to changes in water chemistry, despite the importance of environmental tolerance and biological interactions that may only be visible by examining the community as a whole. In this study, we examined the ability of pH and several water chemistry covariables to explain temporal changes in the littoral benthic macroinvertebrate communities of lakes recovering from acidification. Data from 17 lakes sampled from 1988 to 2002 were summarized using correspondence analysis and compared using Procrustes analysis. Canonical correspondence analysis was used to examine the relationship between chemical variables and community structure. Benthic community composition changed over the sampling period, with significant year-to-year changes from 1993 to 1998. Community composition and water chemistry were highly correlated throughout the study period, although the strongest correlations were found from 1993 to 1997, coinciding with the period of greatest change in the benthic community. These results suggest that benthic macroinvertebrate communities in these lakes have changed in response to changes in water chemistry that are consistent with recovery from acidification.
Live-bait fishery areas off California and Baja-California (modified from Rodríguez-Sánchez et al. 2001), including squares of 2° × 2° from which time series of sea surface temperature were obtained from the COADS database. Areas: (A) California waters; (B) Mexican border to Punta Eugenia; (C) Punta Eugenia to Cabo San Lazaro; (D) Bahia Magdalena – Bahia Almejas – Bahia Santa Maria; (E) Punta Redonda to Cabo Falso; (F) Gulf of California West (Cabo Falso – Bahia de La Paz).  
(a) Geographical variability of smoothed sea surface temperature (SST) anomalies by fishing area along California and Baja California for the 1930-1996 period. Area A is off California and area E is near the tip of the Baja-California peninsula. (b) Time series of yearly (solid line) and smoothed (shaded area) SST anomalies for the full coastal area.
Simplified diagram illustrating the differences between a homogeneous spread resulting from the simple expansion– contraction model of range changes with population abundance increase–decrease and the spatial process described here for the Sardinops caeruleus population in the California Current System. (a) Expansion–contraction model, with the implicit assumption of geostationary population, usually underlying conventional analyses of long data series. (b) A nongeostationary dynamic model proposed for Pacific sardine on a regime scale. In both models, A 1 represents the refuge area occupied when the population size is at a low level, A 2 is that when the population size is growing or diminishing, and A 3 is the largest extension occupied when the population size is at a high level.  
Geographical variability of Pacific sardine (Sardinops caeruleus) abundance along California and Baja California for the 67- year period. The annual catch-per-unit-effort (CPUE) values are smoothed. Area A is off California and area F is in the mouth of the Gulf of California. (a) A three-dimensional view and (b) its projection in two dimensions.  
Geographical variability of northern anchovy (Engraulis mordax) abundance along California and Baja California for the 67- year period. The annual catch-per-unit-effort (CPUE) values are smoothed. Area A is off California and area F is in the mouth of the Gulf of California. (a) A three-dimensional view and (b) its projection in two dimensions.  
Climate regime shifts in the Northeast Pacific appear to have forced population size changes associated with major geographical variations in the position of the center of distribution and bulk of biomass of Pacific sardine (Sardinops caeruleus). These findings help explain the disappearance of sardines around 60 years ago at the northern part of the California Current System and their return following the 1980s. The spatial processes described here differ from those suggesting that environmental regime shifts lead to progressive increase-decrease of stock abundance associated with homogeneous expansion-contraction of its distribution range around a fixed geographical center. Sardine population changes are seemingly related to environmental variability, whereas the spatial pattern of abundance for the northern anchovy (Engraulis mordax) appears to be inversely related to sardine population abundance. Anchovies increased where and when sardines were either absent or at a low population level. Thus, from the long-term and large-scale perspective, neither sardine nor anchovy populations conform to the simple homogeneous geographical range changes usually assumed. The sardine population changes its location within the ocean habitat in an evolving progression over a multidecadal time scale.
Herein, we document changes in the Lake Michigan food web between 1970 and 2000 and identify the factors responsible for these changes. Control of sea lamprey (Petromyzon marinus) and alewife (Alosa pseudoharengus) populations in Lake Michigan, beginning in the 1950s and 1960s, had profound effects on the food web. Recoveries of lake whitefish (Coregonus clupeaformis) and burbot (Lota lota) populations, as well as the buildup of salmonine populations, were attributable, at least in part, to sea lamprey control. Based on our analyses, predation by salmonines was primarily responsible for the reduction in alewife abundance during the 1970s and early 1980s. In turn, the decrease in alewife abundance likely contributed to recoveries of deepwater sculpin (Myoxocephalus thompsoni), yellow perch (Perca flavescens), and burbot populations during the 1970s and 1980s. Decrease in the abundance of all three dominant benthic macroinvertebrate groups, including Diporeia, oligochaetes, and sphaeriids, during the 1980s in nearshore waters (≤50 m deep) of Lake Michigan, was attributable to a decrease in primary production linked to a decline in phosphorus loadings. Continued decrease in Diporeia abundance during the 1990s was associated with the zebra mussel (Dreissena polymorpha) invasion, but specific mechanisms for zebra mussels affecting Diporeia abundance remain unidentified.
To investigate the hypothesis that the 1976 "regime shift" in North Pacific fish populations resulted from climatic change propagating up the fisheries food web, we have embedded a four-component planktonic ecosystem model in an ocean general circulation model. The Miami isopycnic model (MICOM) has been implemented on a 2 degrees grid over the domain from 18 degreesS to 61 degreesN, with a Kraus-Turner-type mixed layer model overlaying 10 isopycnal layers. An initial baseline run with forcing for the period 1952-1988 reasonably reproduces the spatial patterns and seasonal changes in SeaWiFS images. Estimates of annual net and export production compare well with contemporary observations of primary and export production at Ocean Station Papa in the subarctic North Pacific but are low by a factor of 8-10 at station ALOHA near Hawaii. Two subsequent runs with forcing for the periods 1952-1975 and 1977-1988 show the main gyres to strengthen after 1976 with large areas of increased mixed layer depth. In the light-limited subarctic, limited areas of shallower spring mixed layer produced increased phytoplankton biomass, whereas in the nutrient-limited subtropical gyre, increased nutrients (or migration of the subarctic front and the equatorial current system into the gyre) after 1976 correlated with increased plankton biomass.
Main predation flows (t·km-2 ·year-1 ) on shrimp (mainly Pandalus borealis), redfish (Sebastes spp.), and small and large cod (Gadus morhua) estimated by the normal (upper values), seal solutions (middle values in parentheses), and misreporting solution (lower values in square brackets). Note that, for each organism, the predation flows represented (arrows) contribute at least 78% of overall predation mortality. For comparisons, fishing mortality fluxes (t·km-2 ·year-1 ) are also included. Underlined values have a coefficient of variation >75%. Size of arrows indicates the strength of the flow.
Main mortality causes (predation and natural mortality other than predation, i.e., other causes) on small cod (Gadus morhua) estimated by each solution. Normal solution shown in solid bars (total: 0.75 ± 0.13 t·km-2 ·year-1 ), seal solution in open bars (total: 0.80 ± 0.10 t·km-2 ·year-1 ), and misreporting solution in shaded bars (total: 0.79 ± 0.10 t·km-2 ·year-1 ).
Main mortality causes (fishing, predation, and natural mortality other than predation, i.e., other causes) on large cod (Gadus morhua) estimated by each solution. Normal solution shown in solid bars (total: 1.65 ± 0.10 t·km-2 ·year-1 ), seal solution in open bars (total: 1.67 ± 0.11 t·km-2 ·year-1 ), and misreporting solution in shaded bars (total: 1.75 ± 0.29 t·km-2 ·year-1 ).
Main predation flows (t·km –2 ·year –1 ) on piscivorous small pelagic feeders (mainly Atlantic mackerel (Scomber scombrus)), capelin (Mallotus villosus), and sand lance (Ammodytes spp.) estimated by the normal (upper values), seal solutions (middle values in parentheses ), and misreporting solution (lower values in square brackets). Note that, for each organism, the predation flows represented (arrows) contribute at least 79% of overall predation mortality. For comparisons, fishing mortality fluxes (t·km –2 ·year –1 ) are also included . Underlined values have a coefficient of variation >75%. Size of arrows indicates the strength of the flow.  
Mass-balance models using inverse methodology have been constructed for the northern Gulf of St. Lawrence ecosystem in the mid-1980s, before the groundfish collapse. The results highlight the effects of the major mortality sources (fishing, predation, and other sources of mortality) on the fish and invertebrate communities. Main predators of fish were large cod (Gadus morhua) followed by redfish (Sebastes spp.), capelin (Mallotus villosus), and fisheries. Large cod were the most important predator of small cod, with cannibalism accounting for at least 44% of the mortality of small cod. The main predators of large cod were harp (Phoca groenlandica) and grey (Halichoerus grypus) seals. However, predation represented only 2% of total mortality on large cod. Mortality other than predation dominated the mortality processes at 52% of the total, while the fishery represented 46%. Tests were performed to identify possible sources of this unexplained mortality. The only way to significantly reduce unexplained mortality on large cod in the model was to increase landings of large cod above those reported. This suggests that fishing mortality was substantially underestimated in the mid-1980s, just before the demise of a cod stock that historically was the second largest in the northwest Atlantic.
The spatial scales at which the dynamics of subpopulations are synchronized affect regional population persistence and reflect the operational spatial scales of factors regulating populations. I examined phase synchrony in the fluctuations of age-0 winter flounder (Pseudopleuronectes americanus) production among 19 southern New England (SNE), USA, coastal nurseries. From 1990 to 2004, nursery production was synchronized at scales up to similar to 200 km based on spatial trends in cross-correlations of first-differenced age-0 abundance time series. However, sliding window analysis of 1975-2005 time series collected in six nurseries <= 55 km apart in northern SNE indicated that synchrony increased from low to high values in the early 1990s. Synchrony in production also increased among three nurseries <= 65 km apart in southern SNE from 1984 to 2004. Thus, interannual fluctuations in nursery production appeared to become synchronized at coarser spatial scales throughout SNE during the 1990s. This coarsening of the spatial scale of control of winter flounder nursery production was coincident with an increase in the frequency of springs with warm temperatures believed to negatively affect early life history processes. Spatial synchronization of winter flounder nursery ground production could destabilize the age-class structure and population dynamics in the region.
Postseason commercial fisheries data are used to estimate a catch per unit effort (CPUE) – soak time relationship for the 1991–1993 and 1996–1997 Bristol Bay red king crab (Paralithodes camtschaticus) fishery in order to gain regulatory and in-season management insight. Use of commercial fishery data allows our model to capture the influence of biological and environmental effects as well as behavioral responses of crabbers to changing natural and regulatory conditions on CPUE. However, data deficiencies present a variety of estimation challenges, especially when the data are derived from neither a contemporaneous nor a scientifically designed sample of the fleet. A statistical framework for dealing with such challenges is illustrated in this paper. This research uncovered three major results. First, data pooling guided by recursive estimation/hypothesis testing is shown to be essential. Second, the analysis provides insight into CPUE response to changing conditions, whether biological, ecological, or policy induced. Third, it is apparent that more complete and contemporaneous collection of commercial fisheries data is critical to refine the estimation of CPUE - soak time relationships. Then, it may be possible to isolate the inter- and intra-seasonal influence of biological and environmental effects as well as behavioral responses of crabbers to changing natural and regulatory conditions.Des statistiques de la pêche commerciale au Crabe royal (Paralithodes camtschaticus) compilées après la saison de pêche à Bristol Bay en 1991–1993 et 1996–1997 ont servi à estimer la relation entre la capture par unité d'effort de pêche (CPUE) et la durée de mouillage de l'équipement, afin d'obtenir une meilleure évaluation de la réglementation et de la gestion pendant la saison. Les données de pêche commerciale permettent au modèle de capter l'influence sur la CPUE des facteurs biologiques et environnementaux, ainsi que des réactions comportementales des pêcheurs aux conditions changeantes du milieu naturel et de la réglementation. Cependant, des carences dans les données causent plusieurs problèmes d'estimation, particulièrement lorsque les données proviennent d'un échantillonnage de la flotte qui n'est ni concomitant, ni scientifique; nous faisons la démonstration d'un cadre statistique qui permet de résoudre de tels problèmes. Trois conclusions principales s'imposent. D'abord, il est essentiel que le regroupement des données soit fait à la suite d'estimations récursives et d'épreuves d'hypothèses. Ensuite, l'analyse permet de percevoir les effets sur la CPUE de changements de conditions, qu'ils soient de nature biologique ou écologique ou qu'ils soient reliés à la politique de réglementation de la pêche. Finalement, il ressort que la récolte de données plus complètes et plus concomitantes sur la pêche commerciale est essentielle pour améliorer l'estimation de la relation entre la CPUE et la durée de mouillage. Il sera alors peut-être possible de séparer les composantes intrasaisonnières et intersaisonnières des effets des facteurs biologiques et environnementaux et des réactions comportementales des pêcheurs aux changements du milieu naturel et de la réglementation de la pêche.[Traduit par la Rédaction]
Growth trajectories during larval to early juvenile stages in Japanese sardine (Sardinops melanostictus) were backcalculated based on the widths of otolith daily increments from 1996 to 2003 in the nursery grounds, Kuroshio-Oyashio transitional waters. Planktonic larvae hatched near Kuroshio have been transported northeastward from the Kuroshio waters to the transitional waters. We show that the somatic growth rates sharply increased after metamorphosis to the early juvenile stage, whereas previously, others showed that growth rate decreases during the larval stage. Otolith growth rates were more variable during the juvenile stage than the larval stage. Interannual variations in the otolith growth rate after metamorphosis explained more than 80% of variation in abundances of recruits (ca. 8-10 months old) in the Pacific coastal waters, whereas no correlation was found between the growth rate during the early larval stage and abundance. Our result was consistent with the hypothesis of growth rate-dependent recruitment success in the Kuroshio-Oyashio transitional waters.
Guay et al. (2000.Can.J.Fish.Aquat.Sci.57:2065–2075) report a test of numerical habitat models (NHMs) that combine two-dimensional hydrodynamic modeling of the depth and velocity fields of a stream reach with simple biological models. The hydrodynamic model was tested with field data from randomly chosen points in the stream, and the overall models were tested with observations of fish in a "verification" section of the stream. This is a proper procedure, but the execution and interpretation of the test seem flawed, largely by a misunderstanding of the spatial scales appropriate. The NHMs used by Guay et al. make predictions for patches or "tiles" at a scale of 1–25 m2; the hydrodynamic component of the NHMs needs to be tested at the same spatial scale.Guay et al. (2000.J.Can.Sci.Halieut.Aquat.57:2065–2075) présentent un test pour éprouver des modèles numériques de l'habitat (NHM) qui combinent une modélisation hydrodynamique bi-dimensionnelle des champs de profondeur et de vitesse du courant dans une section de cours d'eau avec des modèles biologiques simples. Le modèle hydrodynamique a été éprouvé à l'aide de données de terrain provenant de points choisis aléatoirement dans le cours d'eau et les modèles généraux l'ont été au moyen d'observations sur les poissons faites dans une section de « vérification » du cours d'eau. C'est là une procédure adéquate, mais l'exécution et l'interprétation du test semblent défectueuses, en grande partie à cause d'une incompréhension des échelles spatiales appropriées. Les NHM utilisés par Guay et al. génèrent des prédictions pour des parcelles ou « tuiles » à une échelle de 1–25 m2; la composante hydrodynamique des modèles NHM doit donc être testée à la même échelle spatiale. [Traduit par la Rédaction]
Results from stepwise regression analyses of the relationship between annual mean biomass of each functional group and predictor variables (pH-corrected residual species richness and select limnological variables) during the acidification and recovery periods of the Lake 302S experiment.
Anthropogenic stressors are the Current drivers of loss of global biodiversity and deterioration of ecosystem function (e.g., primary production). However, it is debatable whether human stressors or associated changes in biodiversity better predict the impairment of ecosystem function. Variation in plankton communities during a whole-lake experiment (Lake 302S, Experimental Lakes Area, Canada) was examined to test whether the stressor treatment effect or subsequent stressor-independent variation in species richness best explained interannual variation in aggregate functional properties, Such as productivity or net total biomass. Although significant "biodiversity-ecosystem function" relationships were detected, these correlations were confounded by the negative effect of experimental acidification on species richness. The stressor effect was removed by plotting functional properties against the residuals from the species richness-pH regressions, which generated either negative or nonsignificant relationships. The lack of significant stressor-independent positive relationships between functional properties and species richness highlights the potential greater importance of other mediating factors, such as interactions among multiple stressors, species identity, and altered trophic interactions, at the whole-ecosystem scale.
We previously estimated the predicted effective population size for the endangered winter-run chinook salmon, Oncorhynchus tshawytscha, based on a number of assumptions, including random survival and return of released fish. Here we present data from actual returning spawners, identified to family by microsatellite loci, and calculate the observed effective population size. In 1994 and 1995, the observed effective population sizes were 93.6 and 78.2% of predicted values, respectively, suggesting that the numbers of returning fish were very close to random expectations in 1994 and less close to random in 1995. The ratio of the effective population size to the adult number, N-e/N, was greater than unity for 1994 and approximately 0.5 in 1995. The high ratio in 1994 reflects the success of the breeding protocol to equalize individual contributions and near random returns, while the lower number in 1995 appears to be the result of both less successful equalization and less close to random returns in that year. These findings provide an optimistic outlook for the success of this supplementation program and suggest that the overall effective population size has not been greatly reduced, since returning spawners represent a broad sample of parents and not fish from only a few families.
Reintroduction of functionally important species is considered a key strategy for restoring damaged ecosystems. However, the sudden reappearance of an extirpated species may have adverse ecological impacts, degrading ecosystem services. Therefore, we experimentally reintroduced the crayfish Orconectes virilis into a biologically recovering, formerly acidified Boreal Shield lake (Lake 302S, Experimental Lakes Area, Canada) to determine its effect on the littoral food web following a 17-year absence. In June 2004, a single-factor experimental design consisting of two treatment levels (crayfish-less control versus 1.8 introduced crayfish . m(-2)) was replicated five times for a total of 10 littoral cages (4 m(2)). Orconectes virilis significantly (P < 0.05) Suppressed the total biomass of other benthic invertebrates by 70% primarily because of declines in larval damselflies and midges. In addition, crayfish reduced periphytic biomass by 90% (P < 0.001). Stable isotopic analyses of the mesocosm food webs further indicated that O. virilis likely functioned as an omnivore, exerting direct and possibly indirect effects on other invertebrates and periphyton. Our findings highlight how the reintroduction of O. virilis must be balanced with adequate fish predation to prevent this species from becoming an invader and negatively affecting the productive capacity of boreal lakes.
Baltic Sea herring (Clupea harengus) have been shown to exhibit morphological differences across the marked salinity and temperature gradients in the region. Here we analyse genetic (nine microsatellite loci), morpho metric (skull shape), and meristic (pectoral fin rays and number of vertebrae) variations across seven samples of spawning herring collected from four spawning locations in the Baltic Sea to examine whether morphological variation correlates with genetic and (or) environmental factors. Results suggest that herring is adapting to its environment through a combination of selection and plastic responses. Skull shape, including and excluding size variation, differed significantly among samples, both temporally and spatially. Genetic and morphometric distances were correlated, especially when size variation was excluded from the analysis. When size variation was included, skull shape variation was more closely correlated with environmental distances among spawning locations. Vertebrate number differed among samples and was correlated with environmental distances, whereas the number of fin rays was not. Genetic and geographic distances among samples were not correlated.
Pinto abalone (Haliotis kamtschatkana kamtschatkana) index stations in the San Juan Archipelago were systematically monitored by the Washington Department of Fish and Wildlife from 1992 through 2006. During this period, abalone abundance declined by 77% and the mean shell length (SL) increased 10.4 mm. Abalone densities at all index stations are currently well below the threshold of 0.15 abalone.m(-2) required for successful fertilization. From 1992 to 1996, 16% of individuals encountered measured < 90 mm SL, while only 6% of the individuals from 2003 to 2006 were in this small size class. Similarly, the number of those > 114 mm SL was greater in the 2000s than in the 1990s. The mean SL of all live abalone observed in the 1990s (107.62 +/- 0.87 mm) was significantly different from the mean SL of empty shells (114.21 +/- 2.1 mm), but no difference was detected between the mean SLs of empty shells and live abalone in the 2000s (114.97 +/- 1.42 mm). Taken together, these data suggest recruitment failure from an Allee response to low population densities.
Observed red abalone (Haliotis rufescens) length at time t versus the length at time t + Dt. Lines are the Lt * Lt points; solid circles are the Lt * Lt+Dt points. (a) 1978–1979; (b) 1979–1980; (c) 1980–1981; (d) 1981–1982; (e) 1982–1984.  
Residuals plots when model M1 is used with prior stated in Table 1 and the text: (a) 1978–1979; (b) 1979–1980; (c) 1980–1981; (d) 1981–1982; and (e) 1982–1984.  
Posterior probability density functions of hyperparameters when model M1 is used under three situations corresponding to Tables 13 (S1, S2, and S3): (a) L 1 ; (b) s 2 L1 ; (c) K; and (d) s 2 K. S1, continuous line; S2, dotted line; S3, dotted-dashed line.
Many marine species exhibit temporal variation in individual growth. Yearly variation in growth has been identified for red abalone (Haliotis rufescens) in southern California, USA, but has not been previously incorporated into growth models. In this study, Bayesian hierarchical models were developed to describe variability in growth rates for the Johnsons Lee red abalone population. Although the Bayesian hierarchical modeling estimates are close to estimates of the nonhierarchical highly parameterized model that assigns an estimate of parameters to each data period when the sample sizes are high, the hyperparameters in the hierarchical model are more useful in incorporating the temporal variability into the stock assessment. By ignoring temporal variability, confidence intervals of the estimates of growth can be unrealisti-cally narrow, possibly leading to bias when these models are used for developing biological reference points such as F 0.1,F max, or F x%. The use of a Bayesian hierarchical approach is generally suggested for future growth modeling and for per-recruitment models that include growth when determining precautionary management decisions.
Conservation hatcheries for anadromous salmonids that aim to increase production and minimizing genetic, ecological, and demographic risks have not been experimentally tested for their ability to increase number of adults spawning in the natural environment. The conservation hatchery program for steelhead (i.e., sea-run rainbow trout, Oncorhynchus mykiss) evaluated in this study caused an increase in the number of redds in the supplemented Hamma Hamma River compared with the presupplementation period. Three control populations (nonsupplemented) either remained stable or declined over the same period. The increase in redds from hatchery-produced spawners did not reduce the redd production from natural-origin spawners. The strategy of rearing and releasing adult steelhead accounted for the greatest proportion of redd abundance increases. Environmentally induced differences in spawn timing between the adult release group and anadromous adults of hatchery and natural origin may explain why the adult release group and anadromous adults assortatively formed pairing combinations on the spawning grounds. Although captively reared adults produced the majority of redds in years they were released in substantial numbers, uncertainty regarding the relative reproductive success of this strategy suggests caution in recommending one strategy over the other. A demographic boost to the naturally spawning population was effected while managing to minimize negative ecological consequences.
We describe a Bayesian modelling approach to estimate abundance and biomass of pelagic fishes from gillnet catches in data-limited situations. By making a number of simple assumptions, we use fish sustained swimming speed to calculate the effective area fished by a gillnet in a specified soak time to estimate abundance ( from the number of fish caught. We used catch data from various sampling methods in northern Australia and elicited anecdotal information from experts to build a size distribution of the true population to compensate for size classes that were unlikely to be represented in the catch because of size selectivity of the gear. Our final abundance estimates for various-sized scombrids (0.04-4.17 and istiophorids (0.004-0.005 were similar to what has been estimated for similar species in more data-rich situations in tropical regions of the Pacific Ocean. The model is particularly useful in data-limited situations in which abundance or biomass estimates are required for pelagic fish species of low economic importance. These data are often required for ecosystem models (e.g., Ecopath) that are increasingly being considered as potential tools for ecosystem-based fisheries management.
Current distribution of lake sturgeon Acipenser fulvescens throughout Ontario. Blue shaded areas = waterbodies where Lake Sturgeon currently exist without consideration of population levels. Area beneath black line = Great Lakes drainage; area beneath red line = Winnipeg River drainage; area north of both lines = Hudson Bay drainage. The 22 study sites across Ontario are identified by the conglomeration of grey dots (each dot represents a netting location)  
Black bars = length frequency of Lake Sturgeon sampled in large mesh; gray bars = extra-large mesh  
Loading plot of the first two principal components explaining 63.6% of the variation  
Lake sturgeon (Acipenser fulvescens) stocks are well below historical levels across their natural range. In this study, we examine why lake sturgeon have not substantially recovered to historical levels in a large regulated river (Ottawa River, Canada). Three primary anthropogenic stressors have been identified as potentially limiting lake sturgeon populations in the Ottawa River: (i) commercial harvest, (ii) contaminants, and (iii) water power management. Hypotheses i and iii were tested by comparing lake sturgeon abundance and examining growth among reaches differing in level of commercial harvest and water management regime; hypothesis ii was tested by assessing contaminant loads in lake sturgeon and examining effects on growth and condition. Relative abundance, growth, mortality, and mean size of lake sturgeon did not differ among river reaches with (n = 6) and without (n = 3) a commercial harvest. Mercury was the only contaminant that was elevated. Neither growth nor condition showed any detectable relationship with mercury body burden. Relative abundance of lake sturgeon was greater in unimpounded than impounded reaches; additionally, there is evidence of faster growth in the impounded versus unimpounded reaches, suggesting density-dependent compensation. Water power management appears to be the primary factor affecting lake sturgeon in this river.
Population and community indicators for the impact of fishing are often estimated using abundance estimates instead of raw sampling observations. Methods are presented for testing null hypotheses of nonsignificant impacts and, where possible, for calculating the statistical power. The indicators considered concern populations (intrinsic growth rate, total mortality, exploitation rate, and a new indicator, the change in fishing mortality required to reverse population growth) and communities (k- and partial-dominance curves, a biodiversity index, size spectrum, and proportions of various population groups). The performance of these indicators is compared for the Celtic Sea groundfish community based on achieved precision, statistical power, and availability and estimation method of reference points. Among population indicators, mean length of catch was most precisely estimated and the corresponding hypothesis tests had consistently large powers. Total mortality performed reasonably well. In contrast, both the intrinsic population growth rate and the exploitation rate gave unreliable results. All tested community indicators performed similarly well. Indicators for which the direction of change caused by fishing is predictable, such as the proportion of noncommercial species or piscivores in the community, are promising indicators at the community level.
Long-term variation in Atlantic salmon (Salmo salar) stocks was analyzed in two Barents Sea rivers, the Teno and Naatamojoki, that represent the northernmost distribution area of the species. In contrast to most of the North Atlantic area, these rivers are among a group of northern salmon rivers that, despite wide annual variation in catches, demonstrate no consistent trend for declining abundance. Variations in abundance were generally synchronous for the total catch and numbers of 1-sea-winter (1SW) and 2SW salmon during period of 1972-2003. Part of the variation observed in catches could be related to ocean climate conditions as the mean seawater temperature in July during the year of smoltification for the Kola section of the Barents Sea was significantly related to numbers of 1SW, 2SW, and 3SW salmon in the large River Teno. In contrast, NAO (North Atlantic Oscillation) indices were not related to salmon catches. The latest increase (1999-2001) in salmon catches in these rivers reflects both temporarily improved oceanic conditions and past management measures affecting offshore, coastal, and river fisheries.
Map of Torres Strait, Australia, showing locations of sites sampled during the annual rock lobster (Panulirus ornatus) population survey in 2006. The area of the circles is proportional to the numbers of lobsters caught per transect.
Analysis of deviance for the generalized linear model (GLM) fitted to the real logbook data of the Torres Strait rock lobster (Panulirus ornatus) fishery.
Density histogram of residuals from the generalized linear model (GLM) fitted to commercial rock lobster (Panulirus ornatus) catch and effort data for the years 1994-2006.
Plot of the relationship between the fishery-independent, survey-based abundance indices and the generalized linear model (GLM)-standardized catch per unit effort (CPUE) for the Torres Strait rock lobster (Panulirus ornatus) fishery. The broken line indicates the proportionality between the CPUE-based and surveybased indices.
Torres Strait rock lobster (Panulirus ornatus) abundance indices extracted from generalized linear model (GLM) standardization; circles indicate values estimated using the commercial logbook data of the lobster fishery; crosses indicate values estimated using simulation data that were modified from the logbook data by discounting the fishing effort as a result of fishing power creep.
Commercial catch-per-unit-effort (CPUE) data are often standardized to construct indices of stock abundance. The value of such standardization lies in the improvement in the proportionality between the derived index and true abundance. Using the Torres Strait rock lobster (Panulirus ornatus) fishery in Australia as an example, we first standardized the commercial CPUE data using a generalized linear model (GLM) and then fitted observation error models to the resulting abundance indices and independent abundance data (as estimated by research diver surveys) to examine the proportionality. While the GLM standardization greatly improved proportionality in comparison with the nonstandardized commercial catch rates, it could produce biased results if the model did not explicitly incorporate variables that had caused changes in fishing efficiency. As most catch-effort standardizations do not model the fishing power component simultaneously, this result may serve as a warning to the potential bias in stock abundance indices extracted from GLMs that are underfitted.
Estimated coefficient of variation (CV) versus number of sampled stream reaches for Lincoln–Petersen bias-adjusted population size estimators for bull trout (Salvelinus confluentus) >120 mm (a), >220 mm (b), and >370 mm (c) in the South Fork Walla Walla River, Oregon, 2004. CVs were estimated from the mean of 10 000 simulations for unstratified reaches (solid circles) and from variance-weighted means for the stratified reaches (open circles). Reaches were stratified by river elevation, which was classified as upper, middle, and lower.  
Descriptive statistics for bull trout (Salvelinus con- fluentus) marked (capture counts), counted in snorkel surveys (snorkel counts), and resighted during snorkel surveys (marked fish only) from 2004 field data in the South Fork Walla Walla River, Oregon.
Estimated coefficient of variation (CV) versus number of unstratified sampled stream reaches for >220 mm bull trout using a Lincoln–Petersen bias-adjusted estimator (solid triangles), capture counts (solid circles), and snorkel counts (open circles). CVs are estimated from the mean of 10 000 simulations. Data are from South Fork Walla Walla River, Oregon, 2004.  
Probability of detecting 75% (solid circles), 50% (open circles), and 25% (solid triangles) declines in population size over 30 years. Results are based on the mean of 10 000 simulations for unstratified stream reach samples using a Lincoln-Petersen biasadjusted population size estimator (a), capture counts (b), and snorkel counts (c). Data are from South Fork Walla Walla River, Oregon, 2004.
A comparison of effort (technician hours) and coefficient of variation (CV) for snorkel counts (grey line and grey circles, respectively) and Lincoln-Petersen estimates (black line and black circles, respectively) when sampling between 10 and 39 reaches in the South Fork Walla Walla River, Oregon, 2004.
Using empirical field data for bull trout (Salvelinus confluentus), we evaluated the trade-off between power and sampling effort-cost using Monte Carlo simulations of commonly collected mark-recapture-resight and count data, and we estimated the power to detect changes in abundance across different time intervals. We also evaluated the effects of monitoring different components of a population and stratification methods on the precision of each method. Our results illustrate substantial variability in the relative precision, cost, and information gained from each approach. While grouping estimates by age or stage class substantially increased the precision of estimates, spatial stratification of sampling units resulted in limited increases in precision. Although mark-resight methods allowed for estimates of abundance versus indices of abundance, our results suggest snorkel surveys may be a more affordable monitoring approach across large spatial scales. Detecting a 25% decline in abundance after 5 years was not possible, regardless of technique (power = 0.80), without high sampling effort (48% of study site). Detecting a 25% decline was possible after 15 years, but still required high sampling efforts. Our results suggest detecting moderate changes in abundance of freshwater salmonids requires considerable resource and temporal commitments and highlight the difficulties of using abundance measures for monitoring bull trout populations.
Published data for aquatic bacterial abundance and production in benthic and planktonic environments were collected from the literature to describe relationships between sample means and variances, to explore the factors that affect these relationships, and to estimate the number of samples needed to detect specified differences among means with adequate power. Between 75 and 94% of sample log10(variance) was explained by log10(mean) for both bacterial abundance and production. Differences in mean-variance relationships of bacterial abundance and production due to habitat (river, lake, marine), quantification method, and experimental manipulation (planktonic bacteria) or substrate type (benthic bacteria) were negligible (less than 11% of residual variance from regressions explained). Between 12 and 69 replicates are necessary to detect a 20% difference in means for bacterial abundance and production with a power of 80%. Given the median rate of replication of 3 to 4, the majority of published studies reviewed here are, at best, able to detect differences in means of 50% (planktonic bacterial abundance) or 100% (planktonic production and benthic abundance and production) with 80% power. If effect sizes less than these values are deemed biologically meaningful, then future studies will have to increase sampling effort to enable detection of such differences.Des données sur l'abondance et la production de bactéries benthiques et planctoniques ont été récoltées dans la littérature et analysées pour décrire la relation entre la moyenne et la variance d'échantillonnage, pour examiner les facteurs affectant cette relation, et pour estimer le nombre d'échantillons requis pour détecter une différence entre deux moyennes avec une puissance statistique donnée. Entre 75 % et 94 % de la variabilité du log10 variance peut être attribuée au log10 de la moyenne pour l'abondance et la production des bactéries. Les différences dans les relations entre la moyenne et la variance de l'abondance et la production bactérienne attribuées aux habitats (rivières, lacs, océans), la méthode de quantification et les manipulations expérimentales (bactéries planctoniques) ou le type de substrat (bactéries benthiques) sont négligeables et expliquent moins de 11 % de la variabilité. Les analyses de puissance indiquent que entre 12 et 69 échantillons sont nécessaires pour détecter une différence de 20 % entre deux moyennes d'abondance ou de production bactérienne avec une puissance de 80 %. Puisque la médiane du nombre d'échantillons dans les études utilisées dans cette analyse se situe entre 3 et 4, la majorité de ces études pouvaient au mieux détecter une différence entre les moyennes de 50 % (pour l'abondance de bactéries planctoniques) ou de 100 % (production bactérienne planctonique ou abondance et production bactérienne benthique) avec une puissance de 80%. Il semble donc que si des tailles d'effet plus petites que 50 % sont jugées écologiquement signifiantes, alors les études futures devront utiliser un nombre plus élevé d'échantillons pour les détecter statistiquement.[Traduit par la Rédaction]
We used reported commercial catch data and historical information regarding unreported catches to estimate the abundance of winter steelhead, Oncorhynchus mykiss, in Puget Sound rivers in 1895, the year in which the peak commercial catch of steelhead occurred. We employed a Bayesian analysis to address the uncertainties associated with the estimation process and report abundance estimates for four large northern Puget Sound rivers and for the remaining aggregate of rivers and streams in Puget Sound. The central 90% of the posterior distribution of total abundance ranged from 485 000 to 930 000, with a mode of 622 000. Compared with the 25-year average abundance for all of Puget Sound of 22 000 for the 1980-2004 period, our results show that current abundance is likely only 1%-4% of what it was prior to the turn of the 20th century. Loss of freshwater habitat alone can account for this reduction in abundance only if there was an extraordinary decline in productivity. Our estimates of historical abundance should better inform the development of recovery goals for Puget Sound steelhead.
The freshwater amphipod Diporeia is a crucial part of the food web in the Laurentian Great Lakes, but has faced serious declines correlated with the invasion of zebra mussels (Dreissena polymorpha), except in Lake Superior, which has seen an increase in Diporeia abundance. Speculation on the mechanisms causing changes in Diporeia densities has not included the possibility of evolutionarily distinct lineages of Diporeia within the Great Lakes. In this study, we use cytochrome c oxidase subunit I (COI) DNA sequence data to investigate the evolutionary history of Lake Superior Diporeia relative to the other Great Lakes and consider potential population structuring within Lake Superior based upon depth or geography. Our analyses reveal that Lake Superior Diporeia represent a distinct lineage that diverged from populations of the other lakes at least several hundred thousand years ago. F statistics show that two localities within Lake Superior were significantly differentiated from all other locales, but analysis of molecular variance did not find significant structure based on depth or geography. Genetic diversity within Lake Superior was not correlated with depth, although abundance was significantly negatively correlated with increasing depth.
Relative abundance indices derived from nominal catch-per-unit-effort (CPUE) data are a principle source of information for the majority of stock assessments. A particular problem with formulating such abundance indices for pelagic species such as tuna is the interpretation of CPUE data from fleets that have changed distribution over time. In this research, spatial population dynamics are simulated to test the historical pattern of fishing effort as a basis for making inferences about relative abundance. A number of age-structured, spatially disaggregated population dynamics models are described for both Atlantic yellowfin tuna (Thunnus albacares) and bigeye tuna (Thunnus obesus) to account for uncertainty in spatial distribution and movement. These models are used to evaluate the reliability of standardization methods and a commonly applied model selection criterion, Akaike's information criterion (AIC). The simulations demonstrate the pitfalls of aggregating CPUE data over spatial areas and highlight the need for data imputation. Simulations support simpler models than those selected using AIC for extracting reliable indices of relative abundance.
Hierarchical model configurations and the number of estimated parameters used in this study.
A simulation study to assess the performance of the modified Cormack–Jolly–Seber model to different levels of tag loss and nonreporting. Panels (a) and (b) assume that all tags are retained and all tagged fish are reported on capture when φ is 1.0 and 0.8, respectively. Panel (c) illustrates different scenarios of 50% reporting and 20% tag loss. p (solid circles), capture probability ; φ (open circles), survival probability; Λ, tag retention; Ψ, reporting rate. N (solid, inverted triangles), population estimate.  
A Cormack-Jolly-Seber model was developed to estimate abundance, survival, and probability of capture for juvenile (<1.8 m total length, TL) and adult (>1.8 m TL) raggedtooth sharks (Carcharias taurus) along the east coast of South Africa. Model estimates were adjusted to account for tag loss, nonreporting, and after release mortality. The model was constructed using mark-recapture data from the Oceanographic Research Institute and Port Elizabeth Museum cooperative tagging programs from 1984 to 2004. The adjusted estimate for juvenile survival was 0.56 and that for adult survival was 0.89. The adjusted estimate of probability of capture for juvenile sharks ranged from 0.06 to 0.17, while that for adult sharks was from 0.02 to 0.04. The mean annual abundance of juvenile sharks was 6800 (coefficient of variation, CV = 13%) and adult sharks 16 700 (CV = 9%). The accumulated effect of tag loss, nonreporting, and after release mortality were to reduce the overall estimate of juvenile and adult abundance by approximately 50%. The adjusted estimate of population size for both juvenile and adult sharks over the last decade appears to have remained constant (P > 0.05).
(a) Schematic plan view of the experimental channel used to observe the behavioural response of fall Chinook salmon (Oncorhynchus tshawytscha) smolt to accelerating flows under three different flow conditions. (b) Side view of the opening panel with an opening height of 7.7 cm. (c) Side view of the downstream barrier containing a small opening that allowed the fish to escape the experimental flume. Barrier heights were 0 cm (no barrier), 25.8 cm, and 45.8 cm creating high-, medium-, and low-flow conditions, respectively.  
Fish have evolved intrinsic flight responses, allowing pre-emptive avoidance of potentially threatening situations. To direct downstream migrant fish away from deleterious conditions at darns and other barriers, mechanical devices such as travelling screens and fish bypass systems are often installed. However, field observations suggest that if these structures create areas of rapidly accelerating flow, they do not effectively guide the fish. We studied the avoidance behaviour of actively migrating fail Chinook salmon (Oncorhynchus tshawytscha) smolts in controlled experiments of low-, medium-, and high-flow accelerations. We Measured the response velocity (V-R) and the velocity gradient (VG) over body length. Although V-R varied significantly with flow conditions and increased with increasing water temperature, the median VG at the instant at which smolts displayed,in avoidance response was similar over the range of accelerating flows tested. Results from this study present the first quantitative information about the avoidance behaviour of fish to flow acceleration and should provide data needed to help engineers and biologists develop effective systems to alleviate anthropogenically altered flow regimes. Furthermore, the devised experimental setup provides a Valuable means to test the effects of accelerating flow oil my downstream migrant fish species.
In lake trout (Salvelinus namaycush) lakes of northeastern Ontario, Canada, aerial surveys of fishing activity on individual lakes (N = 589) and quantitative gillnet surveys (N = 65) were used to assess the effects of road access on angling effort and the presence of introduced smallmouth bass (Micropterus dolomieu). Angling effort, particularly during the open-water season. was highest and often exceeded estimated Sustainable levels on lakes with good road access. Approximately 25% of the remote lakes also received excessive pressure during the winter season. Angler numerical responses to lake trout abundance were detected in remote lakes. but 1101 in road-accessible lakes. Smallmouth bass were more prevalent in lakes with road access and human settlement (either cottages or lodges), supporting the theory that they were introduced into these lakes. Lake trout populations were depleted throughout much of the Study range. Even without road access or smallmouth bass, lake trout abundance was still 47% lower than in unexploited reference lakes. When bass and (or) road access were present, lake trout abundance decreased by 77%. Remote lake trout populations in this area are clearly vulnerable to the negative impacts of improved access, a vector for both overexploitation and species introductions.
Pacific salmon (Oncorhynchus spp.) have repeatedly exploited new habitat following glacial recession and some artificial introductions, yet the initial process of colonization is poorly understood. Landsburg Diversion Dam on the Cedar River, Washington, excluded salmon from 33 km of habitat for over a century until it was modified to allow passage in 2003. Adult coho salmon (Oncorhynchus kisutch) were sampled as they entered the newly accessible habitat in the first 3 years and a subset received radio transmitters to assess spawning site selection and movement. Annual counts of coho colonists increased over time, and in 2 of 3 years, daily dam passage was positively correlated with river discharge. Contrary to our prediction that coho would spawn in tributaries, all identified spawning sites were in the mainstem Cedar River, though 38% of radio-tagged salmon entered a tributary at least temporarily. Females moved little within the new habitat (average = 5.8 km), whereas males moved extensively (average = 34.8 km), especially when females were scarce. The immediate use of the new habitat by colonists and their widespread movements suggest that exploration is an innate component of salmon breeding behavior, and restoring access to lost habitat merits priori-tization as a conservation strategy.
Results of repeated-measures ANOVA to test for the effects of hypoxia on weakfish behavioral attributes. 
Percentage of initial swimming speed (at 7.0 mg O 2 ÁL –1 ) that individual saturation-acclimated (*) and hypoxia-acclimated (*) weakfish attained during the DO recovery period. Points show the mean response over the previous 30 min period. The broken line indicates DO exposure.  
Areas of low flushing and high nutrient loading in shallow estuaries are susceptible to diel-cycling hypoxia and also represent important nursery habitat for juvenile estuary-dependent fishes. Swimming speed, angular correlation, and expected displacement were measured in juvenile weakfish (Cynoscion regalis) in response to diel-cycling hypoxia (cycling between 7.0 and 0.4 mg O(2).L(-1)). Saturation-acclimated (7.0 mg O(2).L(-1)) weakfish exhibited an active response by increasing swimming speed (to a maximum at 2.8 gm O(2).L(-1)) and angular correlation (to a maximum at 1.4 mg O(2).L(-1)) as dissolved oxygen (DO) decreased, after which weakfish exhibited a passive response and both swimming speed and angular correlation decreased by similar to 50% and 70%, respectively, at 0.4 mg O(2).L(-1). Weakfish acclimated to hypoxia (cycling between 2.0 and 11.0 mg O(2).L(-1) for 10 days) did not vary swimming speed during decreasing DO or DO recovery (increasing DO) and had an overall swimming speed 46% lower than saturation-acclimated weakfish at 7.0 mg O(2).L(-1). At the end of DO recovery, saturation- and hypoxia-acclimated weakfish had recovered 60% and 80% of their initial swimming speeds, respectively. The relationship between previous hypoxia exposure and behaviour may be an important determinant of habitat utilization in estuarine nursery areas impacted by diel-cycling hypoxia.
Relationship between U MSY and a and under Beverton-Holt recruitment for a hypothetical species. The asymptotic upper limit of U MSY is represented by a broken vertical line. The area to the right of this line represents undefined hypotheses for UMSY.
Mean and modal values of 100 Monte Carlo estimates of U MSY Lim for the 12 species of dogshark.
Deepwater dogsharks (Order Squaliformes) are thought to be particularly Vulnerable to overfishing due to life history strategies that place them at the lower end of the shark productivity spectrum. Sharks frequently have relatively low value in multispecies fisheries, where management is usually aimed at maintaining harvest of more valuable and productive teleosts, This results in low priority being given to data collection for sharks and hampers identification of appropriate harvest strategies. Here an age-structured model with maximum sustainable harvest rate (U-MSY) as leading productivity parameter is systematically applied to show that for certain growth and reproductive schedules that apply to some sharks, the range of possible values that can be taken by U-MSY can become very small. The model was applied to 12 Australian dogshark species and was used to show that U-MSY is highly constrained under some selectivity schedules. Results were consistent with estimates of the intrinsic rate of growth obtained using a demographic model, suggesting that there may be more certainty about U-MSY than expected for many shark species, given uncertainty in recruitment parameters. The approach could be used to inform policy for some sharks and may be useful in the development of informative Bayesian priors for assessment models.
The purpose of this work is to present a conceptual model for fish otolith growth in which somatic growth is related to otolith growth and opacity. The model is based on known mechanisms of CaCO3 and protein incorporation into the otolith. Model parameters were derived from laboratory experiments and from the literature. A sensitivity analysis showed that the model was robust to measurement errors in most input parameters. The most sensitive parameters were the exponents of the otolith length-weight and otolith protein - whole-body protein synthesis relationships and the proportionality constant between metabolism and otolith growth rate. Application to experimental data resulted in good agreement between back-calculated and observed fish sizes. In the growth experiment, the average back-calculated weights were slightly lower than the average observed weights, but the correlation was highly significant. In the starvation experiment, the back-calculated weights were also highly correlated with observed weights, with slightly declining residuals with fish size. Unlike previous back-calculation methods, this model has the ability to detect periods of starvation and estimate growth histories in both growing and starving fish.
We used field data from transplantation and caging Studies With juvenile yellow perch (Perca flavescens) to test a kinetic bioaccumulation model for cadmium (Cd). The model, which considers both dietary and aqueous sources of Cd, was first used to predict the dynamics of Cd accumulation in perch exposed to high ambient Cd for 70 days. Model simulations for hepatic Cd agreed well with the observed time course of Cd accumulation in (lie liver, but for the gills and gut, the predicted accumulations after 70 days were about three times higher than the observed values, suggesting that these latter organs can alter their ability to take Lip and (or) eliminate Cd. The model was also used to predict steady-state Cd concentrations in the gills, gut, and liver of perch living in lakes along a Cd gradient. Agreement between predicted and observed steady-state Cd concentrations was reasonable in lakes with low to moderate Cd concentrations, but in lakes with high dissolved Cd (>1.5 nmol.L-1), the model overestimated Cd accumulation, particularly in the gills and gut. These results suggest that kinetic bioaccumulation models may better apply to some organs than to others. Because metal-induced toxicity is normally organ-specific, their application in a risk assessment context should be undertaken with caution.
Estimating the accuracy of genetic stock identification (GSI) that can be expected given a previously collected baseline requires simulation. The conventional method involves repeatedly simulating mixtures by resampling from the baseline, simulating new baselines by resampling from the baseline, and analyzing the simulated mixtures with the simulated baselines. We show that this overestimates the predicted accuracy of GSI. The bias is profound for closely related populations and increases as more genetic data (loci and (or) alleles) are added to the analysis. We develop a new method based on leave-one-out cross validation and show that it yields essentially unbiased estimates of GSI accuracy. Applying both our method and the conventional method to a coastwide baseline of 166 Chinook salmon (Oncorhynchus tshawytscha) populations shows that the conventional method provides severely biased predictions of accuracy for some individual populations. The bias for reporting units (aggregations of closely related populations) is moderate, but still present.
Bioenergetics studies of free-living animals have long been hampered by limitations on our abilities to measure the energy costs of different activities. Here we evaluate whether it is possible to use the opercular ventilatory beat rate of a fish to estimate its rate of energy expenditure. Changes in metabolic rate (MR) and ventilation rate (VR) were recorded in yearling Atlantic salmon (Salmo salar, weight range 1.8-12.64 g) engaged in different activities at different temperatures while within a respirometer. MR was found to correlate strongly with VR in all fish. The relationship was linear, and both the slope and corresponding intercept of the regression equation were strongly dependent on the fish's body weight and the test temperature. From these relationships, a general equation was generated to predict MR at a range of temperatures from knowledge of a fish's weight and its VR; this proved to be highly accurate (correlation between predicted and observed MRs: r = 0.95), although calibration of individual fish is recommended in studies that compare performance of individuals. Visual measurements of VR may therefore provide a highly accurate, cheap, and noninvasive method of measuring the energy consumption of fish engaged in natural behaviours in more natural settings.
Location of the study lakes in central and eastern Ontario, Canada (45.538N, 82.418W to 44.158N, 76.258W). Sediment was collected from 202 lakes using a modified KB corer. Top (present-day) and bottom (preindustrial) sediment slices were analyzed for total Hg content.
ANOVA of Hg concentrations among various stages of measurement.
Observed and predicted Hg concentrations in (a and b) historical (>30 cm depth) and (c and d) present-day (0.5-1 cm) sediments of 54 headwater lakes standardized for organic matter content. Standardized Hg concentrations were predicted (a and c) before and (b and d) after correcting for autocorrelation using a cubic regression model, where x = latitude and y = longitude. Hghistorical = 0.011 ± 0.002(Ad /Ao) + 0.0008 ± 0.0003(MAR); Hgpresent day = 0.0029 ± 0.0043(Ad /Ao) + 0.0005 ± 0.0006(MAR).
Partial residuals in Hg enrichment from deep-core (>35 cm depth) to near-surface sediments (0.5-1 cm) in 54 headwater lakes as a function of pH after removing the variance attributable to drainage ratio and sulfate concentrations. Hg EF = (Ad /Ao)-15.96-0.07(SO 4
Increase in Hg concentration from deep-core (>30 cm depth) to surface sediments (0.5-1 cm) in 54 headwater lakes of south-central Ontario presented before (*) and after (*) correcting for both pH and sulfate concentration as a function of drainage ratio. Hg EF = (A d / A o )-15.96-0.07(SO 4 2)-(3.55(pH >8.3 )). The mean increase in Hg concentration over the study area is 2.46 ± 1.64 (broken line).
Mercury (Hg) concentrations in recent (0.5-1 cm) and preindustrial (>30 cm) sediments were examined across lakes in south-central and eastern Ontario. Canada (45.53 degrees N, 82.41 W to 44.15 degrees N, 76.25 W). to determine whether Hg exported from watersheds is at steady state with atmospheric deposition. An examination of headwater lakes revealed that Hg enrichment was not uniform among watesheds but that the enrichment factor (EF = [Hg](present day)/[Hg](preindustrial). standardized for organic matter) decreased as a function of drainage ratio (A(d)/A(o), watershed(area)/lake(area)). Furthermore, the model fit was improved after accounting for differences in sulfate concentraions and pH among lakes: EF = (A(d)/A(o))(-15.96) - 0.07(SO(4)(2)) - (3.55(pH(<8.3))) (R(2) = 0.458, p = 0.0001). Hg concentraions in preindustrial sediments of headwater lakes showed a positive linear relationship with drainage ratio (partial t = 4.83, p < 0.0001, n = 66) that was strengthened following an adjustment for mean annual runoff (MAR) ([Hg](preindustrial) = 0.011 +/- 0.002(A(d)/A(o)) + 0.0008 +/- 0.0003(MAR) (R(2) = 0.108, F([1.66]) = 8.01, p = 0.006)). Our results suggest that Hg export from watersheds may be currently lagging behind atmospheric Hg deposition, in which case. Hg export would increase into the future, even as Hg deposition from the atmosphere stabilizes.
Regressions of log Hg (mgÁg dry weight (dw) –1 ) versus d 15 N (%) for fish and littoral invertebrates collected from Beaverskin (solid line, shaded circles), North Cranberry (long-dashed line, Â), Pebbleloggitch (short-dashed line, +), and Puzzle (dotted line, ~) lakes at Kejimkujik National Park and National Historic Site, 2006. Slopes from Beaverskin and North Cranberry lakes were significantly different.  
Four study lakes in Kejimkujik National Park and National Historic Site, Nova Scotia, Canada. Inset shows location in Eastern Ca- nada.  
Mercury (Hg) concentrations in fish from acidic lakes (pH < 6.0) are typically elevated above those from near-neutral systems. It is unknown whether high biomagnification rates through the supporting food web can explain elevated Hg concentrations in top predators from low pH lakes. To investigate this, we collected yellow perch (Perca flavescens), brown bullhead (Ameiurus nebulosus), banded killifish (Fundulus diaphanous), golden shiner (Notemigonus crysoleucas), and littoral and pelagic invertebrates from four acidic lakes in Kejimkujik National Park and Historic Site (KNPNHS), Nova Scotia, Canada, and analyzed them for total Hg and methyl Hg (MeHg), and delta C-13 and delta N-15 to determine sources of energy and trophic position, respectively. Mercury biomagnification rates (slopes of log Hg versus delta N-15) varied significantly among the four lakes but did not explain the among-lake differences in perch Hg; these slopes were also within the range published for near-neutral systems. Rather, Hg concentrations in yellow perch (i.e., predatory fish) in KNPNHS were higher in lakes with higher MeHg in lower-trophic-level organisms and suggest that processes influencing Hg uptake at the base of the food web are more important than rates of food web biomagnification for understanding the variation in concentrations of this contaminant among top predators.
Top-cited authors
Kenneth Cummins
  • Humboldt State University
Pierre Pepin
  • Fisheries and Oceans Canada
Steven E. Campana
  • University of Iceland
Julian D. Olden
  • University of Washington Seattle
John D. Neilson
  • Fisheries and Oceans Canada