CPUE standardisation and the construction of indices of stock abundance in a spatially varying fishery using general linear models

Construction of annual indices of stock abundance based on catch and effort data remains central to many fisheries’ assessments. While the use of more advanced statistical methods has helped catch rates to be standardised against many explanatory variables, the changing spatial characteristics of most fisheries data sets provide additional challenges for constructing reliable indices of stock abundance. After reviewing the use of general linear models to construct indices of annual stock abundance, potential biases which can arise due to the unequal and changing nature of the spatial distribution of fishing effort are examined and illustrated through the analysis of simulated data. Finally, some options are suggested for modelling catch rates in unfished strata and for accounting for the uncertainties in the stock and fishery dynamics which arise in the interpretation of spatially varying catch rate data.     

Cluster analysis of linear model coefficients under contiguity constraints for identifying spatial and temporal fishing effort patterns

For fisheries management purposes, it is essential to take into account spatial and seasonal characteristics of fishing activities to allow a reliable assessment of fishing impact on resource. This paper presents a novel technique for describing spatial and temporal patterns in fishing effort. The spatial and seasonal fishing activity patterns of the French trawler fleet in the Celtic Sea during the period 1991–1998 were analysed by modelling fishing effort (fishing time) with generalised linear models. The linear model for fishing effort included fixed effects for both spatial (statistical rectangles) and temporal units (months). In addition, spatial correlations in any given month were modelled by an exponentially decreasing function. Temporal correlations were included using the previous month's fishing effort for a given spatial unit as predictor. A method based on cluster analysis of estimated model coefficients of spatial or temporal fixed effects is proposed for identifying groups of similar spatial and temporal units. A contiguity constraint is imposed in the clustering algorithm, ensuring that only neighbouring spatial units or consecutive temporal units are grouped. The cluster analysis identified 22 spatial and 9 temporal groups. Winter and spring months stood out as being more variable than the remaining months. Spatial groups were of varying size, and generally larger offshore. The proposed method is generic and could for example be used to analyse temporal and spatial patterns in catch or catch rate data.     

Catch rate standardization for yellowfin tuna (Thunnus albacares) in Taiwan's distant-water longline fishery in the Western and Central Pacific Ocean,with consideration of target change

Catch per unit of effort (CPUE) needs to be standardized to remove the effects of factors such as fishing time and location, before it can be used as an index of abundance in fish stock assessments. One of the most substantial effects arises from a change of target species. This is particularly important for the Taiwanese distant-water longline fishery, which has a long history of fishery data from two fleets that target various tuna species across three oceans. We review the development of the Taiwanese distant-water longline fishery and compare five designs for standardizing the catch rate of yellowfin tuna (Thunnus albacares) in the western and central Pacific Ocean, using generalized linear models with lognormal and delta-lognormal error assumptions. Two approaches to address targeting effects were tested: separating fishing fleet data based on observer records, and including four target indicators calculated from catch data. Four statistical regions (relating to major fishing grounds) were treated as a single factor in the first three cases and were treated separately for the last two (one independent run for each region). The last case, which involved independent analyses for each fishing fleet for each region, and using the delta-lognormal approach, was considered to provide the most informative standardized CPUE trends for yellowfin tuna.     

Theories and behavioural drivers underlying fleet dynamics models

In the domain of decision‐support tools for the management of marine fish resources, considerable attention has been paid to the development of models explaining how fish stocks change over space and time. In most models, fishing effort is assumed to be exogenous and determined by factors such as management. Increasingly, there has been a call for bio‐economic models to also account for the dynamics of fishing fleets, recognizing that fishers respond to changing environmental, institutional and economic conditions. A growing literature has sought to explicitly model the endogenous determinants of the capacity of fishing fleets, the intensity of its use and its temporal and spatial allocation across fishing opportunities. We review this literature, focusing on empirical applications of the behavioural models that have been put forward to explain and predict observed fleet dynamics. We find that although economic factors are usually included as a dominant driver in most studies, this is often based on the use of proxy variables for the key economic drivers, for which adequate data are lacking. Also, while many studies acknowledge that social and social–psychological factors play a significant role in explaining observed fishing behaviour, their inclusion in fishing fleet dynamic models is still very limited. Progress in this domain can only be achieved via the development of multidisciplinary research programmes focusing on applied quantitative analysis of the drivers of fishing fleet dynamics.     

Standardising the CPUE for the <Emphasis Type="Italic">Illex argentinus</Emphasis> fishery in the Southwest Atlantic

The catch per unit effort (CPUE) is a widely used index for assessing the abundance of exploited populations in fishery management. To obtain appropriate CPUE values, it is essential to standardise catch-effort data from fisheries. This task is particularly important for squid fisheries because squid generally have a short life-span and are vulnerable to environmental variability, and thus effective fishery management should take such factors into account. In this study, we analysed unit catches of paired vessels operating under similar fishing conditions to calculate their relative fishing power (RFP) in order to standardise the CPUE of the Taiwanese fleet jigging for Illex argentinus in the Southwest Atlantic. To evaluate the appropriateness of the method, we used a logbook dataset covering eleven years (1993–2003), in which 93.5% of the total catch during the period was included. The results indicate that 98.7% of the fishing effort can be standardised according to the estimated RFP. Compared to nominal CPUE, the standardised CPUE values projected an explainable temporal pattern, indicating an increasing trend in abundance from 1995 to 1999 and a subsequent sharp plunge from 1999 to 2003. However, the RFP was not related to apparent physical factors of the vessel, such as gross tonnage or vessel length. Our evaluations suggest that the RFP method is appropriate for standardising the CPUE, so that it can serve as an abundance index that reflects the annual recruitment size of the squid fishery, because the quality of the method can potentially take possible affecting factors into account in order to satisfy the general assumptions of standardisation criteria. However, the effects of varying the settings of parameters should be carefully examined prior to applying this standardisation method to other squid fisheries.     

基于空间相关性的西北太平洋柔鱼CPUE标准化研究

CPUE标准化方法通常都假设名义CPUE之间是相互独立且没有相关性,然而鱼类集群分布通常存在着空间相关性,为此本研究以西北太平洋柔鱼的CPUE标准化为例,采用1999-2012年6-11月中国鱿钓生产数据以及对应的海表面温度和叶绿素浓度的环境数据,将空间相关性加入广义线性模型(general linear model,GLM)中.在空间GLM模型中运用4个距离模型(指数模型、球面模型、线性模型和高斯模型),进行标准GLM模型和4种空间GLM模型的CPUE标准化结果比较.结果发现,4种空间GLM模型均比标准GLM模型的最小信息准则(akaike information criterion,AIC)更小,标准化结果更准确.同时,在4个距离模型中,指数模型的AIC值最小,其CPUE标准化结果最佳.研究表明,在CPUE标准化中,鉴于鱼类集群与分布特性,应该充分考虑空间相关性这一因素.     

Assessing opportunity and relocation costs of marine protected areas using a behavioural model of longline fleet dynamics

Increasing use of spatial management tools in fisheries requires an understanding of fleet response, and in particular to where displaced fishing effort is likely to move. We develop a state‐dependent decision‐making model to address the spatial allocation of effort in an Australian tuna longline fishery. We assume that fishers have an economic objective in deciding where to fish, but that decisions in any period are also influenced by the remaining quota held at the time of the decision. Key features of the model include endogenous price dynamics, a moving stock and a competitive pool of different vessel types operating from different port locations. We utilize this model to illustrate fleet responses to marine reserves and limits on fishing effort. The results illustrate that the model framework provides advantages over statistically based models in that decisions made in response to the imposition of a reserve are not consistent with a proportional reallocation of effort. Rather, the stochastic dynamic model yielded an overall profit level of ～4% higher relative to scenarios with no reserve. Incorporating the opportunity cost of a quota into the model resulted in an optimal utilization of effort, in which effort was concentrated in time periods and locations yielding maximized profit. Under a low level of effort relative to the season length, the model indicated an overall profit level 43% greater than the highest obtained when the same level of effort was applied solely within any given quarter of the season.     

Examining the importance of consistency in multi-vessel trawl survey design based on the U.S. west coast groundfish bottom trawl survey

Generalized linear mixed-effects models can be used to combine bottom trawl data from multiple vessels, each with a different fishing power, into a single time series of relative abundance. However, how important might it be to have a consistent set of vessels and vessel characteristics from year to year given we can model differences in fishing power among vessels? We demonstrate how changes in the suite of fishing vessels performing the survey can affect the results of the data analysis using sablefish catches in the U.S. west coast groundfish bottom trawl survey from 1998 to 2000. The results do not indicate that one must have a consistent set of vessels over time to provide useful data, but rather that there is benefit to consistency even when the survey data are analyzed using advanced statistical models. Further research should be undertaken to quantify these benefits specifically to aid in contracting and bidding of survey vessels.     

Global fishing effort (1950–2010): Trends,gaps, and implications

According to a recent World Bank report, the intensification of global fishing effort and the ensuing depletion of marine fish stocks causes economic losses of 50 billion US dollars annually. Data deficiencies, however, currently hamper analysis of global fishing effort. We analyzed data from the Food and Agricultural Organization of the United Nations (FAO), the EUROPA fishing fleet registry, and peer-reviewed and other publications, to determine the global trends in fishing effort from 1950 to 2006. Our results show that global fishing effort, expressed as total engine power and the number of fishing days in a year (kilowatt days), was roughly constant from 1950 to 1970, and then steadily increased up to the present. Europe dominated global fishing effort, followed by Asia. Projecting current trends suggests that Asia will soon surpass Europe. Trawlers contribute a major fraction of global fishing effort, as do vessels greater than 100 gross registered tons. Current estimates of global fishing effort, the number of vessels, and total vessel tonnage are, however, underestimates given the data gaps that we have identified. Our results are useful in the following ways: (1) they may encourage researchers in academia and government to improve global fishing effort databases; (2) they allow deeper global analyses of the impact of fishing on marine ecosystems; (3) they induce caution in accepting current underestimates of economic losses of global fisheries; and (4) they reinforce calls for a reduction in global fishing effort.     

利用栖息地指数模型预测秘鲁外海茎柔鱼热点区

利用栖息地指数模型准确地预测了秘鲁外海茎柔鱼的热点区。根据2008-2010年1-12月期间我国鱿钓渔船在秘鲁外海的生产数据,结合实时的海表温及海表面高度数据,分别建立以作业次数、单位捕捞努力量渔获量为基础的适应性指数。利用算术平均数模型建立基于海表温和海表面高度的栖息地指数模型,并利用2011年生产及环境数据对栖息地指数模型进行验证。结果显示,以作业次数为基础的适应性指数符合正态分布,而以单位捕捞努力量渔获量为基础的适应性指数显著性检验不显著,因此,只建立以作业次数为基础的模型。结果表明,以作业次数为基础的栖息地指数模型都高估了茎柔鱼热点区的范围,但大体范围基本一致,这说明其能较好地预测茎柔鱼的热点区。     

Estimation of the overall fishing power: A study of the dynamics and fishing strategies of Brittany's industrial fleets

The present paper suggests a method for estimating the fishing power of vessels and for analysing fleet dynamics. The approach is based on quantification of stocks catchability (q), derived from fishing mortality coefficients (F) as calculated by virtual population analysis. Catchabilities for each harvested stock are thus estimated relatively to the fishing effort (f_n) of each vessel, according to the equation: q = (F/f_n). A linear model is then fitted to these catchabilities. The model allows the identification and quantification of trends in average mortality rates per fishing hour for each stock. Under some assumptions, trends are interpreted as variations in the overall fishing power of each fleet. The approach is applied to three industrial and semi-industrial fleets of Brittany (Lorient, Concarneau and Douarnenez) and to the main gadoid stocks they exploit off the west coast of Scotland (ICES area VIa), and in the Celtic Sea (ICES area VIIf,g,h) for Concameau. Results show large variations in fishing power. Particularly, a marked increase trend in the fishing power exerted on saithe (Pollachius virens) is highlighted for the three fleets, over the period 1983–1989. These variations can be explained by the redirection of fishing strategies, which may occur on a large scale. Thus, we show how the collapse of saithe stock at first led the three fleets to intensify the harvesting of saithe, and from 1989 on, to adopt different strategies. The possible causes of the observed dynamics are discussed, as well as their consequences for fisheries management. In particular, the relevance of direct control of fishing effort as a regulatory tool is questioned.     

Voluntary angler logbooks reveal long‐term changes in a lentic pike,Esox lucius,population

Sixty‐two years of voluntarily collected angling logbook data from a large natural Danish lake were used to study variation in pike, Esox lucius L., CPUE (catch per unit effort), expressed as no. of captured pike per boat trip, as an index of stock size. Pike CPUE was positively related to pike release rate by anglers and negatively affected by certain commercial fishers. The stocking of young‐of‐the‐year pike and a fishery‐dependent index of perch, Perca fluviatilis L., abundance (which may be pike prey or predator depending on size) did not correlate with pike CPUE. Analyses of the size distribution of pike, based on sizes of annual record trophy pike captured by anglers, confirmed the negative impact of commercial pike fishing and revealed a positive influence of air temperature. It is concluded that high‐quality angler logbooks that record effort and catch can be a cost‐effective tool to inform lake fisheries management by revealing long‐term population trends. Further, state space modelling, a statistical technique not yet seen in recreational fisheries science, is recommended as a tool to model proxies for population dynamics from angler logbook data.     

What makes fisheries data informative?

Informative data in fisheries stock assessment are those that lead to accurate estimates of abundance and reference points. In practice, the accuracy of estimated abundance is unknown and it is often unclear which features of the data make them informative or uninformative. Neither is it obvious which model assumptions will improve estimation performance, given a particular data set. In this simulation study, 10 hypotheses are addressed using multiple scenarios, estimation models, and reference points. The simulated data scenarios all share the same biological and fleet characteristics, but vary in terms of the fishing history. The estimation models are based on a common statistical catch‐at‐age framework, but estimate different parameters and have different parts of the data available to them. Among the findings is that a ‘one‐way trip’ scenario, where harvest rate gradually increases while abundance decreases, proved no less informative than a contrasted catch history. Models that excluded either abundance index or catch at age performed surprisingly well, compared to models that included both data types. Natural mortality rate, M, was estimated with some reliability when age‐composition data were available from before major catches were removed. Stock‐recruitment steepness, h, was estimated with some reliability when abundance‐index or age‐composition data were available from years of very low abundance. Understanding what makes fisheries data informative or uninformative enables scientists to identify fisheries for which stock assessment models are likely to be biased or imprecise. Managers can also benefit from guidelines on how to distribute funding and manpower among different data collection programmes to gather the most information.     

Fishing effort redistribution in response to area closures

Spatial fishery closures will induce fishing effort to either move to open areas or to cease to fish. When designing a short- or long-term closed area management regime, the expected impact of that closure will depend upon how that effort is redistributed. We present a redistribution model based upon Ideal Free Distributions (IFDs) which is intermediate in complexity between analyses in which effort is distributed uniformly over open areas and models of full fleet dynamics. The IFD models incorporate the fundamentals of the decision process invoked by fishers facing relocation and the ensuing catch rates that result from the addition or removal of effort. Two classes of models were tested: an Availability model where catch rate declines were proportional to abundance; and an Abundance model where abundance declines at an exponential rate with the entry of displaced effort into an area. Results of these models were compared with uniform and proportional redistribution methods. The IFD-based methods included relative cost of relocation, thereby illustrating the importance of both catch rates and movement costs in designing closed area regulations. To demonstrate the methods, hypothetical area closures to United States pelagic longliners in the western Atlantic were examined and the impact of those closures on bycatch rates was evaluated. Guidance for selecting an appropriate model structure for a particular closed area problem is given.     

First implementation of a Gadget model for the analysis of hake in the Mediterranean

An age-length structured model was built for European hake Merluccius merluccius in the central Mediterranean Sea using Gadget. This analytical framework allowed to integrate multiple sources of information, including fisheries-dependent and fisheries-independent data, collected at different scales and aggregation levels. The model includes the two main fisheries targeting hake in the area, the trawl and the gillnet fisheries, and account for differences in their selectivity and effort. Alternative models were used for testing different assumptions on recruitment and growth. The model is then used to predict the main trajectory of the stock during the next years and to evaluate the potential effects of implementing closed areas management scenarios in the hake nurseries as a specific tool to reduce fishing mortality on recruits. The modelling framework presented performed successfully also in a commercial landing data limited context, common for the Mediterranean. Our results provide statistical support for fast growth and multiple recruitment events assumptions. Including both these key features represent an unprecedented improvement of modelling hake population dynamics in the Mediterranean. We found that the reduction in the fishing effort that characterized the fisheries in the study area during the last few years, coupled with fast recovery abilities of the hake stock, has the potentiality to allow a moderate increase of the stock during the next years. Interestingly, our simulations show that the positive effects which might be expected from protecting hake nursery grounds are only marginally related to a reduction in hake recruits fishing mortality. Although our model relies on assumptions and surely represents an over-simplification of the real world, it still contributed to improve our understanding of the temporal dynamics of one of the most valuable fish stocks in the central Mediterranean.     

Factors affecting the fishing power of sardine fishing vessels and implications for the management of the fisheries on Lake Kariba, Zimbabwe

Abstract The factors important in determining fishing power in the Lake Kariba sardine, Limnothrissa miodon (Boulenger, 1906), fishery were identified using data from the major fishing area of the lake, the Kariba basin, for the periods 1980–1982 and 1988–1990. Comparisons were carried out using fishing power calculated from the mean catch of some standard vessels. Three methods were used for identification of the important attributes. Analysis of variance (ANOVA) was carried out for each component to compare the mean fishing power at all levels within each variable. Multiple regression analysis was used to build predictive models and to determine the factors which best predict fishing power. Factor analysis was used to ordinate vessel types and examine any vessel groupings related to fishing power. Factor analysis showed that vessel and net size were the most important variables. Vessels without engines for propulsion, of low value and without radios had less fishing power than the other vessels in the fleet. Vessels from the same company had similar fishing power, suggesting that some unmeasured variable such as quality of management linked to the fishing companies had a significant effect on catch. The factors which were important in determining catch were length of vessel, depth of the net, presence and absence of echo sounder, mobility, the type of light, number and wattage of the underwater lights in the 1988–1990 fishing period. Vessel length and net category were most important in the period 1980–1982. The models accounted for between 37.6 (1981) and 61.2% (1988) of the variation in fishing power. During the development of the fishery, there has been a change from purse seine to lift nets, and increases in the use of echo sounders, engines for the propulsion of vessels and radios, all of which could have increased effective fishing effort.     

A surplus production model with latent truncation and species targeting,with an application to Papua New Guinean fisheries

This paper first reviews traditional bio-economic models of catch–effort equilibrium and later contributions based on augmented and revised specifications. To overcome some of the pitfalls in fisheries analysis, an approach is formulated which accounts for latent truncation in the fishing fleet, species targeting and non-linear long-term relationships among catch, effort and biomass. The procedure is applied to purse seine and longline offshore marine fisheries in Papua New Guinea, where tuna and other fish resources are believed to be under-exploited on the whole, but selective overfishing is reported to take place. Statistical evidence of incidental truncation is weak, with results being sensitive to the selection of variables. Based on regression diagnostics and expected signs/statistical significance of parameter estimates, non-linear surplus production specifications prove to be more suited than original and unrestricted versions of the conventional approach for modelling the dominant (purse seine) fishery in PNG over the period 1979–2007, with both main and secondary target fishing being found not to exceed the maximum sustainable yield. In either case, policy implications of these results should be pondered against underreporting of official fish catches.     

Development of performance indices for the Newfoundland and Labrador snow crab (Chionoecetes opilio) fishery using data from a vessel monitoring system

This study investigates the potential for using data from a vessel monitoring system (VMS) to create indices of commercial fishery performance that may be used in monitoring snow crab resource status. Fishing hours were screened from hourly positional signals to create an index of fishing effort (hours fished) for comparison with that derived from logbooks (number of trap hauls). Similarly, a VMS-based fishing catch per unit of effort (CPUE) index was developed for comparison with CPUE derived from logbooks. Analysis of these indices showed that VMS-based fishing effort and CPUE indices can be interpreted to provide reliable complementary or alternative indices to logbooks for assessment of fishery performance in the Newfoundland and Labrador (NL) snow crab (Chionoecetes opilio) fishery. We also developed a VMS-based index of fishing efficiency and illustrate how it can be applied toward understanding various behaviors and anomalies in the fishery. VMS data may offer other potential applications for snow crab assessment and management. Our approach and methods are applicable to other commercial fishery resources worldwide that are monitored using vessel monitoring systems.