Skip to main content

Managing fisheries to conserve groundfish and benthic invertebrate species diversity

Deliverables

When different research groups in different countries operating on different research vessels are involved in collecting data for a single dataset, then standardisation of data collection, storage and analysis is paramount. For future comparisons of data, subsequent to completion of this project, it is also necessary for replication of the protocols to be adhered to. As a result a very detailed manual covering all aspects of data collection is produced. This detailed methodology manual describes the samplers used, as well as their deployment, for fish and benthic infaunal and epifaunal invertebrates. The processing of each collected sample is detailed, including preservation and storage methods where necessary. The data to be recorded are listed, examples of data recording forms provided and the procedures for data exchange outlined. The forerunner to MAFCONS, EC FAIR-CT95-0817, developed a standardised 2-metre beamtrawl for sampling invertebrates from IBTS survey vessels. This beam trawl was used in Mafcons and is now standard equipment for many different sampling strategies worldwide. MAFCONS have now produced this sampling manual and parts of this sampling protocol are already being used by researchers on other IBTS survey vessels where invertebarate data are collected (using the 2m-beam trawl).
It was hoped that Huston's (1994) Dynamic Equilibrium Model (DEM) which is based on the relationship between disturbance, diversity and productivity, might be used as an algorithm to assess the impacts on diversity (a requirement of many environmental directives) as a result of disturbance (from fishing activity) in areas of differing productivity within the North Sea. However, three explicit tests of the DEM were carried out and in each the data failed to support hypotheses derived from the DEM and it therefore proved to be inappropriate for the task of providing scientific advice in support of an EAFM attempting to meet biodiversity objectives for the north Sea. The reason for this failure is likely to be that the DEM assumes Lokta-Voltera type population dynamics, whereby population growth is directly proportional to the abundance of mature individuals within the population. For many marine fish and benthic species this is an inappropriate assumption because of the "storage" and "lottery" dynamics of their reproductive output, associated with their non- deterministic growth and the stochastic nature of their recruitment. The DEM also describes the diversity of the species as a function of variation in productivity and disturbance, but where non-deterministic growth is occurring within populations it is preferable to use body size as the functional unit rather than the species. As a consequence, a size structured species diversity model has been outlined and developed, and the analyses carried out to date suggest that such a model may provide useful insight for scientists advising an EAFM addressing biodiversity issues.
Two quarter three (Q3) groundfish survey data sets were analysed, the ICES IBTS, which uses a GOV otter trawl, and the Dutch DBTS. A method for estimating catchability at length for each species sampled in the GOV otter trawl is described. Catchability of different species in the GOV trawl varied markedly. Application of catchability coefficients to the IBTS data suggest that the total biomass of the demersal fish community of the North Sea varied between 3 and 7 million tonnes, with peaks occurring in 1999 and 2000. Cluster analysis of the species abundance suggested the presence of three or four distinct community types across the North Sea. The diversity indices applied to the data were Hill's N0 (species richness), N1 (exponential of Shannon-Weiner Index) and N2 (reciprocal of Simpson's Index). For both IBTS and DBTS data sets it was necessary to aggregate approximately 20 trawl samples in order to arrive at diversity statistics that were representative of the actual diversity. However, this runs the risk of confounding (point) and (transect) diversity. Analysis suggested that where the aggregation of 20 samples could be achieved within a search radius of less than 50km then the estimate consisted entirely of α diversity, but where search radius exceeded 50km the estimates included an increasing component of β diversity. Spatial patterns of Hill's N1 and N2 indices varied between IBTS and DBTS, suggesting they were strongly affected by the variable catchability of the two gears. Application of catchability coefficients to IBTS abundance data generated a different spatial pattern explicable on the basis of environmental parameters. This suggested species diversity was highest in the shallower, mixed, more productive southern North Sea. However, species richness and species diversity of the targeted fish component (log2 weight classes 8 and above) was highest in the northern region when catchability was accounted for, but when based on the raw abundance data diversity appeared highest in the south-eastern region. A method for estimating growth productivity of the fish assemblage, based on the von Bertalanffy growth function and weight at length relationships, has been presented. A linear regression method for estimating the required parameter values for each species is described. The maps of biomass density, production density and P/B ratios are, once again seriously affected by the catchability of the gear. Productivity in the prey-fish component of the assemblage (Log2 weight-classes 3 and 4) was highest in the northern North Sea. Once the MAFCONS consortium have made use of these data and been written up into a number of scientific papers, making the results available to a wider audience, the databases containing these data will be made available to the general public via our website.
Huston's Dynamic Equilibrium Model considers the effect of disturbance on species diversity. In order to test this model as a predictor relating fishing activity and ecological consequences it was therefore necessary to have some means of relating fishing activity to the level of mortality generated by the disturbing event. Therefore, indices of disturbance should be based on a quantification of total mortality, including all target and non-target organisms, induced by all fishing activity in a given area over a period of time, including some measure of impact from habitat alteration. In the context of modelling species diversity and disturbance, the assessment of disturbance must be based on the alteration of population mortality caused by the specified perturbation. The components of mortality to target fish and invertebrates caused by fishing include landings, discards and trawl escapees; whereas the mortality of non-target fish and invertebrates must be modelled on data that quantify levels of fishing activity, namely fishing effort data. Because of the fundamental differences in mobility and spatial distribution between fish and benthic invertebrates, two separate but complimentary approaches were made. Consequently a major review was made of the knowledge of ecological disturbance caused by fishing and the final results have been the production of two modelling approaches that utilise the fishing effort data to estimate mortality within fish and benthic invertebrate communities. From these, maps of ecological disturbance to these components were generated for use in the testing of Huston s model. The invertebrate mortality models were based on recent meta-analysis studies to determine "per fishing event" mortality rates for a range of benthic invertebrate fauna. Recent small scale studies of the Dutch beam trawl fleet have suggested that the distribution of fishing activity follows a Poisson distribution, such that tow velocity, tow duration and gear width for each fishing metier can generate a mean fishing frequency in each ICES rectangle from fishing effort data. This can then be coupled with the "per fishing event mortality" to provide the total benthic mortality in each rectangle to be estimated. The fish mortality model utilised swept area estimates combined with estimates of local density, making assumptions about catchability, to determine the number of fish taken in each fishing event. The number of events per rectangle was then estimated from effort statistics and the model generated annual rates of fishing mortality for each species recorded. Estimates of discards for each commercial species were obtained from annual stock assessments and these were then used to raise landings to total catch, with corrections applied to account for non-target species. Such total catches in each ICES rectangle were then converted to "exploitation" rate indices by dividing them by estimates of the abundance of each fish present.
If an ecosystem approach to fisheries management (EAFM) is to successfully address biodiversity issues then the mechanisms relating diversity and fishing activity need to be clearly understood, this in turn requires that processes that structure marine communities must be similarly understood. Community ecology has been the subject of much research, but mostly directed at terrestrial rather than marine systems. This material is reviewed to establish how much of it is relevant to marine communities. The patterns in terrestrial community structure and the processes (eg competition, predation, disturbance, top-down, bottom up control) that give rise to such patterns are reviewed and examples from marine communities are then examined. Some of the 44 differences recorded between marine and terrestrial communities are profound and suggest that the theory explaining variation in structure of terrestrial communities may be irrelevant when applied to marine systems. One particular difference relates to growth patterns, being characterised by deterministic growth in terrestrial species, suggesting that species are the ecological functional units allowing species diversity to be a useful concept with respect to such communities. However, marine species tend to be characterised by non-deterministic growth, resulting in considerable variation in niche use between different sized individuals of the same species. As a result it makes more sense to consider size classes of organisms as the basic ecological functional unit, rather than species. Thus, studies of the species diversity of marine communities must take account of the size structure of the organisms concerned. Marine organisms tend not to display Lokta-Voltera type population dynamics, common in terrestrial organisms where annual per capita birth rates are constant, but show fecundity proportional to body mass, which continues to increase in non-deterministic growth, and consequently per capita fecundity also increases. As a consequence, the continuing increase in body mass of the individuals remaining after mortality has reduced their abundance, may maintain or even increase the potential for population growth. Furthermore, annual recruitment survival varies considerably between years, such that large cohorts can arise from low levels of reproductive stock. Such population dynamics do not suit models such as DEM, and an alternative size-structured, species-interactive model is proposed which could provide the basis for predicting the effects of fishing on marine fish and benthic invertebrate species diversity.
Sampling of the epibenthic invertebrate community was undertaken using a 2m-beam trawl at stations sampled for fish by IBTS and DBTS (Dutch Beam Trawl Survey). Organisms taken in the samples were identified to species, measured and weighed, allowing size-based approaches to be applied to the resulting abundance data to provide estimates of productivity of the epibenthic community at each sample location. The diversity indices applied to the data were Hill's N0 (species richness), N1 (exponential of Shannon-Weiner Index) and N2 (reciprocal of Simpson's Index). The spatial distribution of species tended to be relatively restricted according to a range of different environmental parameters. Cluster analysis suggested two main communities; a northern and southern community, which exist in very different environmental conditions. Strong latitudinal and longitudinal trends were observed in all three indices. Total epibenthic biomass varied considerably, being lower where particle size was less than 200m and consequently productivity tended to also be lower in these muddier habitats. Productivity was positively correlated with bottom water temperature and depth, although this is unsurprising as temperature is one of the parameters in the models used to estimate productivity and temperature is related to depth. All three diversity indices were negatively related to productivity. Considerable variation in spatial patterns of biomass, and productivity were evident between different sized invertebrates. The biomass productivity of the larger epibenthic invertebrates was least in the southern North Sea. Once the MAFCONS consortium have made use of these data and been written up into a number of scientific papers, making the results available to a wider audience, the databases containing these data will be made available to the general public via our website.
Infaunal invertebrates were sampled using a van Veen grab and the community was described in terms of organisms retained within a 1mm mesh sieve. The same productivity models were used as with the epibenthic fauna, but size structuring was limited to estimates of mean individual biomass of each species retained in a range of sieve sizes and all individuals retained in the sieves were identified to one of 73 different taxon groups. Distributions of the key taxon groups tended to be more widely dispersed than the individual epibenthic species, because of the accumulation of several species per group. The highest overall abundance and biomass of infaunal invertebrates were observed in the southern North Sea. Cluster analysis of the taxon group composition revealed two distinct communities, again occupying the northern and southern North Sea. Taxon group richness and diversity tended to be higher in the northern North Sea, whereas infaunal productivity tended to be highest in the southern North Sea, but with some isolated hotspots of productivity located in the north. Once the MAFCONS consortium have made use of these data and been written up into a number of scientific papers, making the results available to a wider audience, the databases containing these data will be made available to the general public via our website.
The original MAFCONS concept of managing fisheries in the North Sea so as to conserve demersal fish and benthic invertebrate species diversity (requirements under OSPAR, 1992, "Convention for the Protection of the Marine Environment of the North East Atlantic"; CBD "The Rio Convention on Biological Diversity"; and "The European Union Marine Strategy Directive") was based on the premise that, through the manipulation of TACs, the spatial distribution of fishing effort could be influenced such that the detrimental effects of fishing on species diversity could be minimised. This initial objective required the ability to predict fishing effort patterns associated with each set of TACs. This relationship remains problematic. However, the data collected during the course of MAFCONS provides an invaluable source of information for identifying areas of particular ecological importance. The impacts of fishing activity can then be minimised in these areas by the use of Marine Protected Areas (MPAs) and thereby assist the EC in its requirements for meeting its ecological goals within the various legal frameworks. Not only can such data suggest appropriate areas for closure but the models developed within MAFCONS can help quantify the fishing effort displaced by an MPA into the surrounding areas in order to maintain landings (financial return) and can also assess the amount by which landings (if TACs still operable) or effort (if effort controls are employed) would need to be reduced in order to prevent further ecological damage to the surrounding areas and can be used for any future Impact Assessment requirements by EC.
A prerequisite in ecosystem management of fisheries is the ability to predict the ecological consequences of management actions. To date fisheries management has involved catch limitations, however the prediction of ecological consequences of fishing are generally modelled on variation in fishing effort. This therefore requires knowledge of the relationship between landings and effort and therefore the relationships between landings, TAC/quotas, and effort were investigated by gear, year and species, in five individual case studies based on partner countries. In each case study, landings and quotas were closely related for the main target species. Relationships between landings by particular gears and the effort by that gear produced confusing results. In each case study there was a strong spatial relationship between landings and effort based on ICES rectangles. However, the strength of temporal correlations was variable across the North Sea. The relationship on a North Sea wide basis between TAC and landings was very close for all fish except saithe, where landings were substantially below the TACs. However, there was no indication that variation in TAC influenced variation in fishing effort in any clear and consistent manner. As a consequence, adequately predicting likely patterns of fishing effort, and thereby likely ecological consequences, from particular combinations of TACs was not possible. This makes continued management through catch limitation difficult to reconcile with a proactive ecosystem approach to management. This fact will be raised with appropriate ICES working groups as well as better ways of recording and accessing fishing effort data. All these data will be incorporated into appropriate scientific publications, making the material available to the wider public.
This database contains files for international fish landings and fishing effort for the North Sea, contributed by all MAFCONS partners participating in their International Bottom Trawl Surveys (IBTS) (England, Wales, Northern Ireland, Scotland, Germany, Netherlands and Norway) including data on vessels from other countries landing into partner countries. The data is accumulated by gear, year (1997 to 2004 incl.) and ICES rectangle (and by species with respect to landings). Data is included for six gear categories: otter trawl directed at fish for human consumption; otter trawl directed at Nephrops; otter trawl directed at other invertebrates; industrial otter trawl; beam trawl; and seine gear. Data is included for seven species: cod, haddock, whiting, saithe, plaice, sole and Nephrops. Once the MAFCONS consortium have made use of these data and been written up into a number of scientific papers, making the results available to a wider audience, the databases containing these data will be made available to the general public via our website.