© 2006 International Council for the Exploration of the Sea
Indicators of the health of the North Sea fish community: identifying reference levels for an ecosystem approach to management
a Fisheries Research Services, Marine Laboratory PO Box 101, Victoria Road, Aberdeen, AB11 9DB, England, UK
b The Centre for Environment, Fisheries and Aquaculture Science, Lowestoft Laboratory Pakefield Road, Lowestoft, Suffolk, NR33 0HT, England, UK
*Correspondence to S. P. R. Greenstreet: tel: +44 1224 876544; fax: +44 1224 295511. e-mail: greenstreet{at}marlab.ac.uk.
The shift in emphasis away from the single-species focus of traditional fisheries management towards an ecosystem approach to management requires application of indicators of ecosystem state. Further, an ecosystem approach to management requires the identification of ecological reference points against which management objectives might be set. In applying indicators, identifying reference points, and setting objectives, an obvious requirement is that the indicators respond primarily to the anthropogenic activity being managed and are sufficiently sensitive that impacts of the activity and the responses to management action are clearly demonstrable. Here we apply a suite of 12 indicators to Scottish August groundfish survey data collected in the northern North Sea over the period 1925–1997. These include indicators of size structure, life-history character composition, species diversity, and trophic structure within the community. Our choice of analytical design has two purposes; first to show that fishing has unequivocally affected these various aspects of the structure of the groundfish community, and second to illustrate an approach by which long time-series data sets might be used to identify possible management reference points. The results are discussed in the context of selecting ecological indicators in support of an ecosystem approach to management and determining appropriate reference points for objective-setting.
Keywords: community metrics, Ecological Quality Objectives, ecosystem approach to management, effects of fishing, fish communities, indicators, reference levels
Received 13 September 2005; accepted 22 December 2005.
 |
Introduction |
|---|
 Top Introduction MethodsResultsDiscussion and conclusionsReferences |
|---|
Recent years have witnessed a shift in emphasis in the underlying objectives of fisheries management in the North Sea. While the need to conserve individual stocks at sustainable and economically viable levels is still clearly a priority, managers are now being asked to adopt a broader ecosystem approach to managing fisheries (Gislason et al., 2000; Sainsbury and Sumaila, 2001; Hall and Mainprize, 2004; Cury and Christensen, 2005; Garcia and Cochrane, 2005). Over the period 1982–2002, five "North Sea Ministerial Conferences" and two "Intermediate Ministerial Meetings" have provided the political impetus for this change in attitude. Commitments to the Convention on Biological Diversity (CBD) and Agenda 21, agreed at the United Nations Conference on Environment and Development (UNCED), and the signing of the Convention for the Protection of the Marine Environment of the North-East Atlantic (known as OSPAR because it effectively combined the two original Oslo and Paris Commissions) strongly influenced these deliberations. In 1995 the Esbjerg Ministerial Conference tasked OSPAR with the responsibility for developing the "ecosystem approach to management", and a series of workshops was convened to meet this requirement. The requirement for objectives at a "general" level was immediately obvious, but the need for "specific" objectives that could serve as detailed operational goals soon became apparent. A basic template consisting of a list of 10 Ecological Quality (EcoQ) Issues, for each of which specific Ecological Quality Objectives (EcoQOs) could be set, was therefore proposed (Lanters et al., 1999). "Fish Communities" is (EcoQ) Issue number 5. The 1997 Bergen Intermediate Ministerial Meeting Statement of Conclusions intimated that EcoQOs for community level EcoQ issues should address multispecies-, assemblage-, or community-wide attributes (e.g. Washington, 1984; Magurran, 1988), initiating a search for appropriate indicators of community "status" or "health" (e.g. Trenkel and Rochet, 2003; Shin et al., 2005; Fulton et al., 2005). This has been particularly intensive with respect to fish communities, because appropriate groundfish survey data are readily available for many maritime areas. In the North Sea, for example, several internationally coordinated surveys have covered the entire area each year for several decades. Numerous different metrics conveying information on a variety of community attributes have been applied to these data (e.g. Greenstreet and Hall, 1996; Rogers et al., 1998, 1999a, b; Greenstreet et al., 1999a; Jennings et al., 1999a, 2002a; Jennings and Reynolds, 2000; Rogers and Ellis, 2000).
Indicator selection
To introduce some order into the process, the International Council for the Exploration of the Sea's (ICES) Working Group on Ecosystem Effects of Fishing Activities (WGECO) proposed a set of criteria to judge the effectiveness of different indices in common use at the time (ICES, 2001; Table 1), and these criteria have started to influence the selection of particular metrics to serve as indicators (e.g. Cury et al., 2005). WGECO concluded that metrics based on the mean size of fish, the proportion of large fish, and the ultimate body size of fish in the community best indicated the impact of fishing on fish communities (ICES, 2001). Ministers at the 2002 North Sea Ministerial Conference in Bergen subsequently adopted these characteristics as the elements of EcoQ for Fish Communities (Bergen Declaration, 2002). A problem with this approach is that these criteria concentrate almost entirely on demonstrating an unequivocal effect of an activity (e.g. fishing) on a particular metric (e.g. mean size of fish in the community). Emphasis is placed on choosing metrics that score highest across all criteria. Little emphasis is placed on the importance of the observed change in the community, whether it is of real ecological significance requiring immediate remedial action, or whether it is trivial and could be ignored by managers with little consequence to the fish community or broader marine ecosystem. A more positive approach might be first to identify the aspects of community structure deemed valuable, or critical to ecosystem function, i.e. to determine user needs, then to select the most appropriate metrics to monitor change in these attributes. This sort of prioritization is advocated by Rice and Rochet (2005). For example, the Convention on Biological Diversity, Agenda 21 and Annex V of OSPAR all place a legal obligation on signatories (including the EC and all European states that border and exploit the biological resources of the North Sea) to conserve biodiversity and to restore it in degraded systems. As no link has been demonstrated between the mean size of fish in fish communities and any one of the three components of biodiversity, within-species genetic diversity, species diversity, and community diversity, knowledge of any change in the size of fish in the community will not help managers implementing an ecosystem approach to management to respond to one of the most important policy drivers imposed on them to date. If the conservation and restoration of biodiversity are management objectives for fish communities, then indicators of biodiversity will be required (e.g. Bundy et al., 2005).
|
Biodiversity is, however, not the only issue on the agenda of an ecosystem approach to management. The various community indices in common use today quantify different aspects of community structure and provide different levels of information regarding the functioning of communities within the broader ecosystem. The relationship between biodiversity and ecosystem function has stimulated much recent debate (Emmerson and Huxham, 2002; Huston and McBride, 2002; Mooney, 2002; Raffaelli et al., 2002). If less diverse communities are less productive and less stable (Tilman and Downing, 1994; Loreau et al., 2002; Petchey et al., 2002; Tilman et al., 2002a, b), then fishing-induced reductions in fish species diversity (e.g. Greenstreet and Hall, 1996; Greenstreet et al., 1999a; Blanchard et al., 2004) may hold wider ecosystem implications beyond simply the effects on the exploited fish community. Community-averaged metrics of life-history traits convey information regarding the types of species present in a community, and these may also suggest changes in community/ecosystem function. In communities where average age at maturity has declined and average growth rates have increased (e.g. Jennings et al., 1999a), increased productivity could be inferred for example. Where assemblage total biomass has remained relatively constant (Yang, 1982; Daan et al., 1990; Sparholt, 1990), but the biomass harvested has increased, such changes might imply increased resource cycling rates within the community. The prospect of fishing down trophic levels in marine foodwebs, altering the trophic structure of fish communities, has also been viewed with some concern (Beddington, 1995; Pauly and Christensen, 1995; Pauly et al., 1998, 2001; Heath, 2005). A holistic ecosystem approach to management that considers all aspects of change in community structure, and the consequences of such to ecosystem function, requires the application of a suite of indicators (Cury and Christensen, 2005; Fulton et al., 2005).
We illustrate this approach here. A suite of 12 metrics is applied to long-term groundfish survey data from the northwestern North Sea in an attempt to identify potential EcoQOs for fish communities. Three potential indicators are applied to examine change in fish size in the community: the percentage of fish exceeding a body length of 30 cm, average weight of fish, and average Lengthinfinity (L
; the ultimate body length in the von Bertalanffy growth function – see Methods) of fish. To address issues related to community productivity, changes in the life-history characteristics of fish in the community are examined. Three possible indicators are applied: the average age and the average length at maturity, and the average growth rate of fish in the community (as indicated by the growth term in the von Bertalanffy growth function – see Methods). To examine trends in fish community biodiversity, five potentially useful metrics are used: the number of species in samples, Margalef's index of species richness, Pielou's index of species evenness, and Hill's (1973) N1 and N2 indices of species diversity. Finally, to explore possible changes in the trophic structure of the fish community, trends in the average Nitrogen stable isotope ratio are examined. When the ICES criteria were applied to these metrics in the past, several scored relatively poorly (ICES, 2001). Demonstrating a clear link between human activity and indicator performance was one of the major difficulties encountered, thus failing criteria b, c, and e (Table 1). Minimizing the impact of fishing would be a primary goal of an ecosystem approach to fisheries management in the North Sea. Therefore, in applying these indices to fish abundance data, an analytical design is adopted such that specific hypotheses related to variation in fishing activity are tested. In this way the causal link between human activity and index performance should be more fully established, and any reservations introduced by the application of these three criteria regarding the usefulness of particular metrics be reduced.
Recent work has led to the proposal of four a priori hypotheses as to how some characteristics of the groundfish species assemblage in the North Sea may be affected by fishing (e.g. Greenstreet and Hall, 1996; Jennings et al., 1998; Greenstreet et al., 1999a; Jennings et al., 1999a), and these underpin the analytical design. In the areas most affected by fishing
- the proportion of large fish, the mean weight of fish, and the average ultimate body length of fish in the groundfish assemblage should be least;
- the species richness and species diversity of the groundfish assemblage should be lower;
- growth rates should be highest, and size and age at maturity should be least; and
- the average trophic level at which fish feed should be lower.
All four predictions follow from first-order effects of fishing as a source of mortality that is not equal across all species and sizes of fish in the community. More complex ecological processes, such as interspecific competition, top-down/bottom-up control, resource supply and cycling, trophic level transfer efficiency, and productivity (Paine, 1974; Connell, 1975, 1978; Carpenter et al., 1987; Huston, 1994; Pauly and Christensen, 1995; Verity and Smetacek, 1996; Pauly et al., 1998, 2001) could all serve to amplify these changes. Previous studies have adopted a pseudo-correlative approach, for example, by determining temporal trends in species diversity (Greenstreet et al., 1999a), or assemblage-averaged growth rate (Jennings et al., 1999a), in areas where fishing effort has increased over time. Whereas such studies might suggest fishing effects on the fish community, the correlative approach used does not definitely link cause and effect. Here we combine spatial and temporal analyses to strengthen the case that the changes observed are indeed caused by fishing.
Reference levels and objectives
Setting actual EcoQOs for community indicators remains a significant obstacle to the use of these metrics within an ecosystem approach to fisheries management. EcoQOs were defined as "the desired level of EcoQ relative to a reference level", where reference levels were defined as "the level of the EcoQ where the anthropogenic influence on the ecological system is minimal". There are thus two parts to the problem: first, establishing reference levels, the metric value expected in the absence of the human activity; and second, deciding a metric level that is consistent with good ecosystem governance, yet still permits the continuance of a viable fishing industry. Addressing the second part of the problem is primarily a political question with significant social implications (Jennings and Dulvy, 2005). However, identification of appropriate reference levels is very much a question that should be addressed by marine scientists. Accordingly, the analytical design adopted here is also intended to address the problem of identifying reference levels for each community metric applied to the fish community of the northwestern North Sea.
|
Methods |
|---|
|
TopIntroductionMethodsResultsDiscussion and conclusionsReferences |
|---|
Analytical design
Data were analysed for 75 ICES statistical rectangles (0.5° latitude by 1° longitude) divided into three treatments of low, medium, and high current fishing effort. The area considered includes nearly half the surface area of the North Sea. A total of 12 potential indicators of different characteristics of the groundfish assemblage was examined to determine the extent to which each was affected by fishing. The hypothesis that each metric was most affected in the rectangles of highest fishing disturbance, and least affected in the rectangles of lowest disturbance, was tested. Even should the data support these initial hypotheses, this does not necessarily confirm that fishing has been responsible for any observed differences. An alternative interpretation, however unlikely, that fishing activity may have been attracted to areas where the groundfish assemblage may have displayed particular characteristics, such as low average weight, or low species diversity, cannot at this point be discounted. To rule out this alternative interpretation, long-term time-series trends for each of the groups of rectangles were examined. If fishing were responsible for the change in the community characteristics, then predictable temporal trends should be apparent. Little or no long-term trend should be apparent in rectangles where fishing disturbance is low, whereas in rectangles affected by fishing, temporal trends in a predictable direction should be detected. The greater the impact from fishing, the steeper should be the gradient (Figure 1). An assumption underlying this analytical design is that prior to any apparent fishing effects, the community characteristics in the different treatment areas had the same start point, and that each was subjected to approximately the same low level of fishing disturbance. This is examined in a later section dealing with reference levels and the setting of target values.
|
In adopting this analytical design, we have attempted to follow, as closely as possible, a one-way ANOVA design. However, it is important to realize that the distribution of fishing effort was not random across the 75 rectangles (Greenstreet et al., 1999b; Jennings et al., 1999b; Figure 2). A true ANOVA design would have had each treatment set of rectangles distributed randomly across the 75 rectangles. This has two major implications. First, spatial variation could introduce a potentially confounding effect. Concentration of the impact of fishing into restricted areas could magnify the effect of fishing on the demersal fish community. Nevertheless, this is still a fishing effect. It will lead to similar distribution in the community characteristic being investigated. The question is whether this spatial factor can introduce the type of trend we anticipate independently of fishing, but this seems unlikely. Second, spatial cohesion of both the treatment and the effect could, through spatial auto-correlation, reduce the independence of individual statistical rectangle data. This has consequences for estimating the actual degrees of freedom in any statistical analysis. While we have presented significance levels for the ANOVA results, some caution is therefore necessary in interpreting these, because "naive" degrees of freedom were used. Analysis of groundfish survey data collected at high spatial resolution (25–30 half-hour GOV samples collected within a 20 x 20-km area) suggests that spatial auto-correlation in species abundance in the catches was entirely diminished at distances of around 10–15 km, i.e. well within the scale of the ICES statistical rectangle (Stransky, 1998). However, examination of semi-variograms for each community characteristic suggested some spatial auto-correlation at distances of up to 150 km (2–3 ICES rectangles).
|
Data sets
Groundfish survey data
Scottish August Groundfish Survey (SAGFS) data collected in 75 ICES statistical rectangles located in the northwestern North Sea where survey coverage was most complete were examined (Figure 2). Only trawl samples collected using a low headline 48-ft Aberdeen Otter Trawl towed for 1 h were included in the data set. Only data for demersal groundfish species likely to be well sampled by this gear were analysed. Pelagic species, such as herring, sprat, and sandeels, were all excluded. The results therefore only apply to the demersal groundfish community occupying the area. For more details regarding the data, see Greenstreet and Hall (1996) and Greenstreet et al. (1999a).
To determine contemporary levels of each community metric for each rectangle, species at length abundance data covering a period of 14 years from 1983 to 1996 were extracted. All samples in this data extraction were collected by the same survey vessel, FRV "Scotia". For one rectangle only 10 trawl samples were available within this time period. This rectangle was not sampled in 1983, 1985, 1987, or 1995. To avoid sample-size dependence problems, sampling effort was standardized to 10 trawls in the other 74 rectangles by excluding, as necessary, trawl samples selected at random from these four years. Previous analysis of SAGFS data has indicated that it is necessary to aggregate at least five 1-h trawl samples in order to derive reliable community metrics, particularly for species richness and diversity indices, which are notoriously sample-size dependent (Southwood, 1978). All 10 trawl samples in each rectangle were therefore combined to provide a single aggregated, highly standardized, species abundance sample for each rectangle upon which to calculate each community metric (henceforward termed the contemporary data subset).
For the second analysis, looking at long-term temporal trends in each of the groundfish assemblage metrics in rectangles varying in the level of fishing effort to which they had been subjected, data from the full time-series, spanning the period 1925–1996, were used. Following Greenstreet et al. (1999a), data were pooled into groups of two or three years to ensure adequate sampling effort in each time-period/effort treatment cell.
Fishing effort data
International fishing effort (hours fished) for the period 1990–1995 were used to define rectangles subject to three different levels of fishing disturbance (Jennings et al., 1999b, 2000). Total annual average fishing effort across the 75 statistical rectangles amounted to 963 216 h of fishing, 67% of which consisted of otter trawling, 12% of beam trawling, and 21% of seine-netting. The otter trawl is the dominant gear used in this part of the North Sea (Greenstreet et al., 1999b). The treatment effort levels set were therefore based solely on that gear. Average annual effort values were calculated to provide estimates of the current spatial distribution of fishing effort across the 75 ICES statistical rectangles for which groundfish survey data were available. These ranged from 645 h year–1 to 63 794 h year–1. Three broad categories (treatments) were defined: 40 rectangles of relatively low otter trawling intensity, from 0 to 4999 h year–1, 25 rectangles of medium otter trawl effort, from 5000 to 19 999 h year–1, and 10 rectangles of high otter trawling intensity, exceeding 20 000 h year–1 (Figure 2).
Greenstreet et al. (1999b) concluded that fishing effort by UK vessels landing into Scottish ports accounted for a large proportion of total international fishing effort in the northwestern North Sea. A more recent report updating this earlier paper directly compared patterns of international effort with those of UK vessels landing into Scottish ports, and provided strong support for this earlier conclusion (Greenstreet et al., 2006). Both studies contend therefore that spatial and temporal trends in fishing effort by UK vessels landing in Scotland are indicative of trends in total international effort in this region of the North Sea. Otter trawl fishing effort by UK vessels landing in Scottish ports increased steadily over the period for which data were available (1960–1998), but the distribution of effort across the region varied little. Across the region the increase in otter trawling effort has been almost, but not quite, matched by a decrease in seine-netting. Beam trawling has only been recorded in the northwestern North Sea since the mid-1980s. As the fish assemblage would only have been subjected to disturbance from beam trawling for the past 10–15 years of the 71-year time-series, beam trawl effort was not taken into account in assigning ICES rectangles to fishing effort treatments. However, examination of the effort data of UK vessels landing in Scotland suggests that inclusion of beam trawl effort would not have influenced the classification of rectangles to effort categories. Examination of the effort data for UK vessels landing in Scotland confirms that rectangles classed as high effort on the basis of the 1990–1995 international effort data, would also have been the most heavily fished during the early 1960s, with a similar situation prevailing for rectangles assigned to the medium and low effort categories. We make the assumption, therefore, that the relative distribution pattern of fishing effort across the region would have persisted back to the start of our data series in the mid-1920s, although we have no means with which to test this.
Community indicators
Size composition metrics
Species abundance at length data in the contemporary standardized data subset were examined to determine an appropriate length at which a fish might be considered to be large. In all, 95% of the individual fish recorded were 
30 cm in length. The top five percentile of the length distribution, fish >30 cm long, was therefore defined as large. Length–weight relationships were used to convert species abundance at length to weight at length. Average weight in a treatment was determined simply by dividing the total weight in the sample (estimated through application of the appropriate species length–weight relationships to numbers at length and summing over all species and lengths) by the total number of individuals. Lengthinfinity (L) data, determined from the von Bertalanffy growth equation calculated for each species, were available for 28 of the 56 species included in the SAGFS database (Table 2; Jennings et al., 1998, 1999a). These 28 species accounted for more than 98% of the individuals sampled by the SAGFS in any of the spatial/temporal treatments. Species abundance data were converted to the number of individuals with particular L values, and the mean value for each spatial/temporal treatment was computed. L could also be considered to be a life-history characteristic and as such, this metric could also be used in this context (e.g. Jennings et al., 1999a). However, in this paper we treat L as an indicator of fish size because of the emphasis placed on the "proportion of large fish in the community" in the Element of Ecological Quality selected for "Fish Communities" in the Ministerial Declaration of the Bergen 2002 North Sea Conference. If average L in the community declines, the potential for the community to contain a high proportion of large fish must also necessarily be reduced.
|
Life-history characteristic metrics
Information regarding three life-history characteristics, growth rate, and age and length at maturity, were available for the same 28 species (Table 2). The growth rate data used were the parameter values determined from the von Bertalanffy growth equation calculated for each species. The von Bertalanffy growth parameter is not strictly a rate value, but is used here as an index equivalent to growth rate. Values of age and length at maturity were determined by observation, either from recent survey data, or with recourse to the literature (Jennings et al., 1998, 1999a). Species abundance data were converted to the number of individuals with particular characteristic values, and the mean value for each characteristic for each spatial/temporal treatment was computed.
Species richness and diversity metrics
Species diversity consists conceptually of two different aspects of species relative abundance; the actual number of species included in any particular sample, and the evenness of the distribution of individuals between the species encountered. Here we use five different indices each differing in the extent to which they are influenced by one or other of these two aspects of species diversity (Southwood, 1978). Species richness was simply the count of all species encountered in the aggregated samples. This index is highly dependent on sample size, so we also apply an index that takes account of the number of individuals included in the sample, Margalef's species richness index, d, determined as (S–1)/ln(N), where S is the number of species and N is the number of individuals in the sample (Clarke and Warwick, 2001). As a straight metric of evenness, we apply Pielou's (1969, 1975) index, J, given as H/ln(S), where H is the Shannon–Weiner index computed for the sample, determined as –
1i pi ln pi, where pi is the proportion of the total sample contributed by the i(th) species (Magurran, 1988), and S is the number of species recorded in the sample. Two diversity indices, Hill's (1973) N1 and N2, were also computed for each of the 75 rectangles' aggregated samples. Hill's N1 diversity index is the exponential of the Shannon–Weiner index, and is therefore e– pi ln pi. N2 is the reciprocal of Simpson's index, D, where D is computed as 1ipi2, where pi is again the proportion of the total sample contributed by the i(th) species (Magurran, 1988). Therefore, N2 is 1/1ipi2. N1 is more sensitive to the number of species recorded in the sample, whereas N2 is more sensitive to the evenness of the distribution of individuals between species.
Trophic level metrics
Increase in the 15N:14N ratio (henceforth referred to as the Nitrogen Ratio) reflects a higher trophic level diet (Minagawa and Wada, 1984). Relationships for Nitrogen Ratio at length were available for 26 of the 56 species encountered in the SAGFS database (Table 3; Jennings et al., 2001, 2002a), accounting for more than 98% of the individuals sampled in any spatial/temporal treatment. Data on species abundance at length were converted to the number of individuals with given Nitrogen Ratio values, and the mean value for each spatial/temporal treatment was computed.
|
15N
and estimated trophic level of North Sea fish. The form of the fitted relationships is |
Results |
|---|
|
TopIntroductionMethodsResultsDiscussion and conclusionsReferences |
|---|
Plots of contemporary mean values of each metric in rectangles of high, medium, and low fishing effort (with confidence limits indicated), along with the results of a one-way ANOVA, are shown in Figure 3. These data were also analysed using non-parametric Kruskal–Wallis analysis, and this gave near identical probability values to the parametric ANOVA. With the exception of the length at maturity (Lmat) and Nitrogen Ratio metrics, for which no obvious effect of fishing was apparent in the contemporary data, all other ANOVAs were significant at p < 0.05 or lower. Except for the two species richness indices, these differences were in the direction anticipated by the stated hypotheses. Both species richness metrics indicated no difference between the high and medium fishing effort treatments, whereas in rectangles with little fishing effort, species richness was significantly lower. Where the results were in the predicted direction, two different effects were observed. Thus, gradients across the fishing effort treatments were apparent in the percentage of large fish, the average fish weight, and the average age at maturity in the community; values were low on high fishing effort rectangles, intermediate on medium effort rectangles, and highest on low fishing effort rectangles. In all other instances where variation in the metrics supported the hypotheses (the three diversity indices and average L
), no difference was detected between high and medium effort treatments, but both differed from the low effort treatments.
|
Long-term trends for the 12 community metrics were determined for each of the three fishing effort treatments. Regression analysis statistics for these trends are provided in Table 4. Our hypotheses impart the following predictions regarding these regression statistics. First, gradients of the slope in the low effort treatment rectangles should be indistinguishable from a slope of zero. Second, gradients of the slopes in medium and high effort treatment rectangles should be statistically different from zero in a predicted direction. Third, the gradients of slopes in a predicted direction should increase as one progresses from low, through medium, to high effort treatment rectangles. Table 5 summarizes the performance of each metric against the three predictions. In some instances, statistically significant outliers from the regression fits were detected (i.e. a significant studentized residual; Velleman and Welsch, 1981). In all cases where this occurred, the regression analysis was repeated excluding these outlying data. In each instance the effect of doing this was to reduce the gradients of the regression lines in high and medium effort treatments, or to reduce the residual variance in the low effort treatments, thereby making it easier to obtain a significant trend in these latter cases. With regard to our hypotheses, therefore, excluding the outlying data was a conservative action, making it more difficult to support the hypothesis concerned.
|
|
For two metrics, significant long-term trends were observed in rectangles with low fishing effort. Thus, even in rectangles where fishing activity was low, average growth rates in the fish community increased over the period 1925–1996, whereas Margalef's index indicated a decline in species richness. Although these significant trends in the low effort treatment rectangles were not anticipated by our hypotheses, the gradients of the trend lines for both metrics were, as predicted, steeper in the medium and high fishing effort treatments. Long-term trends in the Nitrogen Ratio were not significant in any of the three fishing effort treatments. For all other metrics, statistically significant trends in the directions predicted by the hypotheses were observed in the high fishing effort treatments, and in all cases, again as predicted, the gradients of the slopes were steeper than the trend-line gradients fitted to the low effort treatments. Trends in the percentage of fish >30 cm and species richness (count of species) were both not statistically significant in the medium fishing effort treatment. However, for all remaining metrics, and in line with the hypotheses' predictions, significant trends were detected on the medium effort treatments, all with steeper trend-line gradients than those fitted to the low effort treatments.
With two exceptions, therefore, the potential fish community indicators examined here have behaved in the way predicted by the initial hypotheses. Gradients of the long-term trend lines in the low effort rectangles have either not differed significantly from a slope of zero, or where significant trends were detected, the gradients were shallower than those in the high and medium effort treatments. These two exceptions are considered further in the discussion, but for simplification in the next step of the analysis, the two metrics involved, Margaleff's species richness index and the von Bertalanffy growth parameter, are treated as if their long-term trends were zero. These results suggest that the low effort treatments provide an indication of a situation where the anthropogenic influence on the ecological system is minimal. Under these circumstances, the mean and the 95% confidence limits of the low effort treatment data might be considered as possible reference levels against which EcoQOs might be set.
Figure 4 shows these reference levels for all 12 community metrics, and against these, the trend lines for the medium and high effort treatments have been plotted. These plots suggest that the SAGFS data set is sufficiently long-lived to have captured the point in time where fishing activities in the northwestern North Sea started to influence the characteristics of the fish community examined. Figure 5 shows the same reference levels as Figure 4, but here the actual time-series data for the high fishing effort rectangles are shown. These plots suggest that with respect to the fish size metrics, the fish community since 1970 in the most heavily fished parts of the area was outside the reference level lower 95% confidence limit for >95% of the time. An almost identical situation is apparent for the three life-history metrics. At first glance the situation does not seem to be so bad with respect to fish community species richness and species diversity. Since 1970, only some half the data points for the five metrics fall below the reference level lower 95% confidence limit. However, the trend lines shown in Figure 4 and the data in Figure 5 suggest a problem with the original assumptions illustrated in Figure 1. It is clear that species richness and diversity in the high and medium effort rectangles at the start of the SAGFS period were actually higher than those in the low fishing effort rectangles. Therefore, the assumption that the three different fishing-effort-level treatment trend lines have deviated away from a similar start point is violated. It would seem that levels of fishing activity have been greatest in areas of high fish species richness and diversity, and that fishing activity in these areas has reduced fish species richness and diversity to the same, or even lower, levels found in rectangles of low fishing effort. Under these circumstances, the mean metric values in the low effort treatment do not adequately represent the non-anthropogenically influenced situation. The reference level lines for these metrics in Figures 4 and 5 should in fact be higher, in which case the situation since 1970 with respect to species richness and diversity would almost certainly be similar to the fish size and life-history characteristic metrics. With respect to changes in the trophic structure of the fish community in the northwestern North Sea, variation in the Nitrogen Ratio would appear to suggest that this has been unaffected by variation in fishing activity.
|
|
|
Discussion and conclusions |
|---|
|
TopIntroductionMethodsResultsDiscussion and conclusionsReferences |
|---|
Evaluation of metrics as indicators
The analyses presented here provide evidence that 11 of the 12 community metrics indicate a response by the fish community to variation in fishing effort. Thus where particular metrics were previously deemed to be weak in respect of criteria b, c, and e following application of the ICES criteria (Table 1), these shortcomings have now, to some extent, been redressed. The data on which the metrics are based, the numbers and lengths of fish captured in trawl surveys, are easily and accurately measured by trained scientists and technicians on board research vessels. Such groundfish surveys, covering entire marine regions, are regularly carried out, and have been for many decades in the North Sea and in other OSPAR areas. Thus, these metrics should all score highly against criteria d, f, and g. While the concepts represented by some of the metrics are easily understood by non-scientists, for example change in the average size of fish in the community, for other metrics, such as those implying changes in the species diversity of the fish community, the implications may be less obvious (Rochet and Trenkel, 2003). In some cases the user need will require that some of these more obtuse indicators be applied (Rice and Rochet, 2005). For example, managers need to know about changes in fish community species diversity in order to respond to policy drivers such as the CBD and Annex V of OSPAR. So rather than abandoning the use of such metrics as indicators in support of an ecosystem approach to management because they fail some of these criteria, it must remain the job of scientists to ensure that managers understand the messages being conveyed by these more technical, and perhaps more difficult, metrics (Cury et al., 2003).
The theoretical linkage between the behaviour of some of the metrics examined here and variation in fishing activity is thought to be well understood. For example, because fishing operations tend to be strongly size-selective, imparting higher mortality rates on the larger fish in the community and little or no mortality on the smallest fish (Beverton and Holt, 1957), and because size-based tropho-dynamics theory suggests that, with a reduction in the abundance of large predators, predation rates on smaller fish that are their prey should decline (Pope, 1991; Kerr and Dickie, 2001), there is a strong theoretical basis to the belief that fishing disturbance should cause indicators of fish size in the community to show a trend towards smaller size. Because of this apparent theoretical linkage between anthropogenic activity and a metric response, and because metrics of fish size appear to show such strong compliance with the ICES criteria, such metrics have been widely used. However, in a recent review, Shin et al. (2005) suggest that the use of these metrics as indicators is not without some inherent drawbacks. Environmental and density-dependent effects on growth and recruitment may also affect metrics of fish size regardless of the level of fishing activity (Ricker, 1995; Ottersen and Loeng, 2000; Lekve et al., 2002). Poor recruitment may cause the average size of fish in a community to increase as populations become progressively dominated by older fish (Wilderbuer et al., 2002). Conversely, increased rates of recruitment may cause the mean size of fish to decline even in the absence of overexploitation. In situations where overexploitation has been followed by remedial action (e.g. fishery closure), coincidental increases in recruitment may inhibit the anticipated increase in the values of size-based metrics (Badalamenti et al., 2002).
The effect of fishing on species with different life-history characteristics is also well founded in theory. Increasing rates of fishing mortality should provide species with r-strategy life-history characteristics, such as fast growth, small body size, and early age at maturation, with a selective advantage over species with the opposite, K-strategy characteristics of late maturation, slow growth and large ultimate body size (Jennings et al., 1998). This hypothesis has received support in previous studies (Jennings et al., 1999a; Greenstreet and Rogers, 2000), and is again strongly supported here. As predicted, both average age and average length of fish at maturity, and average L were lower in ICES statistical rectangles with higher levels of fishing activity, and long-term temporal trends were more steeply negative, while the reverse was true in respect of average growth rates of fish. The analytical design strongly suggests that these differences and changes in the life-history composition of the fish community were caused by variation in fishing disturbance. Similar effects also occur within species, where high fishing mortality situations effectively influence the evolution of these traits within populations (Law and Grey, 1989; Rijnsdorp, 1993; Conover and Munch, 2002; Grift et al., 2003; Olsen et al., 2005). Unfortunately, data are not available that would allow us to examine the extent to which this is occurring throughout the fish community, but if such within-species responses are widespread, then the trends in life-history characteristics reported here may actually be understating the situation. Such fishing-induced changes in the life-history traits of fish communities may also need to taken into account when considering the use of size-based metrics as indicators. Stobberup et al. (2005) suggest that metrics of fish community size structure may be poor indicators of overexploitation in fish communities characterized by fast growth, small body size, and early age at maturation. Thus, because of the changes in community life-history character composition caused by fishing, as communities become increasingly perturbed by fishing, metrics of fish community size structure may become less effective as indicators of continuing damage.
This discussion serves to make the point that, despite their apparent advantages, metrics of size in fish communities should still not be used casually. The potential for processes other than fishing mortality to influence trends in metric values needs always to be considered. Despite these reservations, in most studies where metrics of size in fish communities have been applied, the anticipated results have been observed (e.g. Zwanenburg, 2000; Bianchi et al., 2000; Blanchard et al., 2005; Daan et al., 2005; Piet and Jennings, 2005). Blanchard et al. (2005) consider possible confounding effects caused by environmental variation, but conclude that fishing had the stronger effect on community size structure. In the current study, the analytical design suggests strongly that variation in fishing effort is the principal cause of differences in the size structure of the groundfish assemblage. Fish communities in ICES statistical rectangles subjected to higher levels of fishing activity had a smaller percentage of large fish, lower average fish weight, and lower average ultimate body length, and long-term temporal trends in these characteristics were more steeply negative, than in rectangles where fishing activity was less.
In aquatic ecosystems, larger predators are generally assumed to feed at higher trophic levels than smaller ones (Kerr and Dickie, 2001; Jennings et al., 2001, 2002b; Jennings and Warr, 2003). Consequently, variation in the size structure of fish communities has generally been assumed to reflect changes in trophic structure within the community (Murawski and Idoine, 1992; Rice and Gislason, 1996; Jennings et al., 2002c). However, when trends in the trophic structure of fish communities have been explicitly examined, results have often been inconclusive or contradictory. For example, in a study of fish communities of the northern and southern Benguela Current, indices of trophic level in the catch remained relatively stable despite the important changes taking place within the system (Cury et al., 2005). In contrast, in the Irish and Celtic Seas, a marked fishing down the foodweb effect was observed, as mean trophic level of the catch declined considerably over the period 1973–2001 (Pinnegar et al., 2002; Gascuel et al., 2005). Application of species- and size-specific stable 15N stable isotope ratio figures to Quarter One ICES International Bottom Trawl Survey species abundance at length data suggested a decline in trophic level within the whole North Sea groundfish assemblage over the period 1980–2001 (Jennings et al., 2002a; Piet and Jennings, 2005). However, when the same analysis was applied to Scottish August Groundfish Survey data covering the northern North Sea only, but collected over a much longer period, 1925–1996, no trend in trophic structure was apparent (Jennings et al., 2002a), leading to the conclusion that changes in the size structure of fish communities caused by fishing may be decoupled from changes in trophic structure. In this study, no significant difference in stable Nitrogen isotope ratio was observed between the three effort treatments in contemporary times, and no long-term trends were detected in either the high or the medium treatments.
Given the strong suspicions that fishing has affected the trophic structure of the North Sea fish community (Heath, 2005), these results are somewhat difficult to accept, and can be interpreted in a number of ways in relation to Rice's (2003) signal theory. Either fishing has not affected fish community trophic structure, in which case this result is a "true negative", or fishing has affected community trophic structure, but this metric is either insensitive to the fishing effect, or is perhaps not a good indicator of trophic structure. In either case, the result should be considered as a "miss". Cury et al. (2005) took this view when their study failed to detect a change in trophic level of the catch in the Benguela Current, suggesting the index was conservative, responding only slowly to large structural changes. If there are serious doubts about the results reported here being a "true negative", then this clearly highlights the need for more research, both to examine trends in the trophic structure of the North Sea fish community and to identify a reliable indicator. Alternatively, if this result is indeed a "true negative", that the trophic structure of the North Sea has remained stable, then this also is a result deserving of follow-up study. When considered in conjunction with the other results presented here, it implies that the larger, less productive, and presumably mostly piscivorous species, which have been lost from the community, have been replaced by smaller, faster growing fish that are also piscivorous. Trophic structure may be ecologically constrained and, so far, robust to the factors causing changes in the fish community. If recycling rates have increased, but the trophic structure has remained unchanged, then this has important implications for ecosystem dynamics, and underlines the need for research into this topic (Jennings and Blanchard, 2004). To close this discussion on trophic level indicators, Bundy et al. (2005) suggest that fisheries that exploit fish evenly across the trophic level spectrum would be less likely to alter the trophic structure of the system than fisheries that concentrate on upper trophic level fish. Since the 1970s, with the expansion of industrial fisheries (ICES, 2004), fishing exploitation in the North Sea might be considered to fall more within the former category than the latter.
Of all the potential metrics of fish community structure that could be used as indicators of fish community health, the use of species diversity and associated indices appears to be steeped in the greatest controversy. Yet given the policy drivers which managers are currently facing, the need for such indicators would seem to be great. Two major difficulties hinder the use of metrics of species diversity as indicators; they are difficult to interpret and the effect of fishing on them is not easily predicted (Rochet and Trenkel, 2003). Consequently they perform poorly against criteria for a good indicator (ICES, 2001).
As the concept of species diversity covers two quite distinct aspects of the distribution of individuals between species, first the number of species and second the evenness of the distribution of individuals between these species, the use of single univariate parameters as metrics of species diversity itself introduces problems (Hurlbert, 1971). Managers informed enough to be aware of this problem but lacking a detailed knowledge of the different metrics in use will therefore have difficulty in understanding exactly how fish communities have changed. The various species diversity metrics available differ in their sensitivity to these two different components of species diversity, from simple counts of the number of species in a sample that are the basis of species richness indices, to indices of evenness that represent the extent to which individuals are evenly distributed across all species in the sample. Hill (1973) noted that a number of different species diversity indices in common use were mathematically related. He ranked these from N0 to Nx such that, with increasing subscript notation, these metrics spanned the range from indices of species richness to indices of species evenness. In reporting metrics of species diversity, it has often been suggested that at least two should be used so as to adequately represent these two quite different aspects of species diversity, giving rise to the notion of using a suite of indices, or "composite indices" (Buckland et al., 2005; Fulton et al., 2005; Wilsey et al., 2005). Most studies reporting changes in fish species diversity in the North Sea have, at the very least, reported variation in two of the first three of Hill's metrics, N0 to N2 (Greenstreet and Hall, 1996; Greenstreet et al., 1999a; Piet and Jennings, 2005). Here we present data for Hill's N0 (Species Richness), N1, and N2, along with a second, more sample-size independent index of species richness, and a further index of species evenness. Taken together, these indices suggest that fishing has caused both a reduction in the species richness of the groundfish community and a less even distribution of individuals between the species present, so that in recent years, the community has become increasingly dominated by a smaller number of more abundant (proportionately) species. Both components of species diversity have in fact declined.
The need for a second, more sample-size independent index of species richness highlights a further problem with these metrics; the fact that metric values are influenced considerably by variation in sample size. This is most clearly exemplified by Island Biogeography theory, wherein it is amply demonstrated that species richness (Hill's N0) increases as a power function of the area sampled, S = cAz, where S is the species richness count and A the area sampled, c is a constant and z the exponent of the sampled area (MacArthur and Wilson, 1967; Rosenzweig, 1995). The implication here is that one can never obtain a true estimate of the actual species richness of a community until the entire community has been sampled; the estimate of species richness will continually increase as an ever increasing area is sampled. Traditional sampling theory is based on the underlying premise that each sample mean provides an estimate of the population mean. This is clearly not the case in this instance. No single (reasonably sized) sample can possibly estimate the species richness of the sampled community; community species richness will always be higher than the sample species count. Calculating the mean species richness value for a number of samples of the community is also clearly a pointless exercise, because the mean value determined for several single samples still remains the estimate of species richness for an area sampled by a single sample, and therefore is just as poor an indicator of the actual species richness of the community as would be obtained from any single sample individually (Colwell et al., 2004). To estimate species richness of a community, successive samples need to be aggregated, such that one can track both the increase in the number of species contained in all samples combined and the increase in the total combined area sampled. In this way, through linear regression of Log A on Log S, it is possible to obtain estimates of c and z and so parameterize the above relationship to determine an estimate of species richness in the community. Such an approach has, for example, been adopted to estimate total species richness of a variety of different plant and animal taxa in many different regions of the world (see examples in Rosenzweig, 1995), but such an approach has not to date been applied to the analysis of groundfish survey data. Because all species diversity indices are sensitive to a greater or lesser extent to increase in the number of species contained in a sample, they will all tend to behave in the same way: as the area sampled increases, so the index value will tend to increase. However, as indices become more sensitive to the evenness of the distribution of individuals between species, and less sensitive to increase in the number of species (i.e. with increase in suffix number in Hill's notation), they become less affected by variation in sampling effort (Soetaert and Heip, 1990).
The ability of different diversity indices actually to detect environmental and anthropogenic impact on communities has been questioned on more than one occasion (Chadwick and Canton, 1984; Robinson and Sandgren, 1984), but generally these problems have been associated with inadequate sample size and failure to understand the importance of determining the required level of sampling beforehand (Soetaert and Heip, 1990). Therefore, in any study of species diversity data, a preliminary analysis of the relationship between index value and variation in sampling effort is a critical first step to determine at what sampling effort level the index values start to stabilize, and so start to represent the true community diversity rather than just being a consequence of the level of sampling effort. Greenstreet and Hall (1996) carried out this preliminary analysis and determined that in order to represent the true diversity of the groundfish community of the northwestern North Sea, a minimum of five 1-h trawl samples needed to be combined. They decided that such a level of sampling was adequate for Hill's N1 and N2 indices, but was still inadequate for Hill's N0 (species richness count). In subsequent studies (Greenstreet et al., 1999a), and including the current study, all treatment cells have included the aggregation of many more than five trawl samples, ensuring that the index values, including the index of species richness, are as representative of actual community values as possible, and as a consequence, these studies have been able to demonstrate trends in diversity associated with variation in fishing impact. Other studies have examined trends in diversity estimated as the mean of single haul samples (e.g. Piet and Jennings, 2005). In the case of the ICES International Bottom Trawl Survey (IBTS), samples consist of single 30-min trawls, one-tenth of the sampling effort deemed necessary in the analysis of Scottish AGFS data, the Scottish precursor to participation in the coordinated Quarter 3 IBTS. Not surprisingly, therefore, these studies have failed to demonstrate fishing effects on fish species diversity. Failure to standardize analytical methodology with respect to the application of species diversity indices to groundfish survey species abundance data has resulted in these inconsistent results, making interpretation difficult and contributing to the general confusion regarding the value of such indices as indicators of fish community health.
Predicting the response of fish community species diversity to variation in fishing disturbance also poses a problem. Again the results of different studies have proven to be contradictory. Thus, while analyses of Scottish AGFS data suggest that fishing has caused declines in fish species diversity in the northern North Sea (Greenstreet and Hall, 1996; Greenstreet et al., 1999a, this study), analysis of various data sets for the southern North Sea suggests that the reverse is true (Rogers and Ellis, 2000; Piet and Jennings, 2005). Once again, however, the problem may well be one of lack of knowledge and understanding. Productivity in the southern North Sea, characterized by shallow, mixed, warmer water, is greater than in the deeper, stratified, cooler northern North Sea (Reid et al., 1990). Huston's (1994) dynamic equilibrium model predicts that species diversity can respond both positively and negatively to increasing levels of disturbance, relationships being positive in areas of high productivity, and negative in areas of low productivity. Therefore, Huston's model actually anticipates these apparently contradictory trends in groundfish species diversity in the North Sea. The actual processes by which the groundfish community of the North Sea are structured, be these top-down (e.g. predation, disturbance mortality) or bottom-up (e.g. resource availability, competition), are ill defined, and for scientists to be able to offer adequate advice on matters concerning species diversity in particular, a much better understanding of the processes is needed. Rather than discard species diversity metrics as possible indicators of fish community health, the policy drivers that managers currently have to address suggest that scientists should instead be addressing these shortcomings.
Determining reference levels and EcoQOs
Current definitions suggest that reference levels for indicators should reflect the situation in the marine ecosystem prior to any changes resulting from anthropogenic activity. By 1900, approximately one million tonnes of fish were being harvested annually by fisheries in the North Sea (Daan et al., 1990), approximately 10% of the fish standing-crop biomass (Yang, 1982; Daan et al., 1990; Sparholt, 1990). This has led many to conclude that the historical data available are unlikely to describe the unexploited state of fish communities. The time-series available are too short, and exploitation has generally preceded scientific data gathering (Jennings and Dulvy, 2005). The data set analysed here represents the longest, more or less continuous time-series available for the North Sea. Even so it only extends back to 1925, so on no account can the data be considered to reflect the unexploited state (Pauly and Maclean, 2003). There can be little doubt that the removal of 10% or more of the fish standing-crop biomass each year for at least 25 years must have affected in some way the structure and composition of the North Sea fish community by 1925. However, a valid question remains: had the community been affected by these levels of fishing activity to such an extent that changes in the metric values would have been detected? If the answer to this question is no, then the values obtained at the start of this time-series do indeed reflect the unexploited state. Unfortunately, however, this must remain a rhetorical question, because there is no data set available with which to answer it. Moreover, variation in the level of North Sea fishing activity is not the only factor capable of influencing fish community structure that has changed in the last 100 years. There have been environmental changes too, and these also can influence the structure of marine communities (Clark and Frid, 2001). Therefore, the unexploited state of a fish community in today's environment may be very different from the unexploited state at the turn of the last century (Jennings and Dulvy, 2005).
Few published studies have speculated on the nature of the unexploited state of fish communities, so currently there is a dearth of information on which to base any realistic assessment of reference levels. One exception to this is the study by Jennings and Blanchard (2004), who applied macro-ecological theory to predict the abundance and size structure of the unexploited North Sea fish community from a theoretical relationship between abundance and body mass. They suggest that the biomass of fish of mass greater than 4 kg is currently >97% lower than would be the case in the absence of fishing. However, this method does have some disadvantages in that the models used apply to all animals present in each size class and take no account of the high species- and size-selectivity inherent in using commercial and scientific fishing trawls. Even scientific trawls fail to quantitatively sample invertebrates and the smaller fish that dominate biomass in the smaller size- or mass-classes (Jennings and Dulvy, 2005). Until such issues can be resolved, the approach adopted in this study may provide at least an interim solution. The data collected at the start of the time-series analysed in this study may not predate the start of fishing in the North Sea, but they certainly predate two major changes in the fishing history of the region: the resumption of widespread fishing activity following the end of World War 2, and the massive expansion of fishing activity that occurred during the 1970s when annual landings increased rapidly to around 3.5 million tonnes (Daan et al., 1990).
Excluding the Nitrogen ratio index of trophic level, for which none of the effects predicted by our initial set of hypotheses were observed, for the remaining 11 metrics our hypotheses were largely upheld. Metric values were either higher or lower depending on the prediction for each metric, and rates of long-term change greater, in ICES statistical rectangles subjected to greater levels of fishing activity. For nine of the metrics the long-term temporal trend was not significantly different from zero in the low effort rectangles. We argue that, for these specific metrics applied to this particular data set, the data might be used as a baseline against which to compare the data in the high and medium effort treatment rectangles. In other words, these low effort treatment data provide the best indication of an undisturbed state and can therefore be used to set management reference levels. For two of the metrics, Margalef's species richness index and the von Bertalanffy growth parameter index of growth rate, significant long-term temporal trends were also detected in the low effort treatments. These may perhaps indicate an environmental effect on the behaviour of these two metrics. Warming water temperatures may, for example, give species with faster growth rates a selective advantage. With increased productivity, species richness may have declined. A fish community increasingly consisting of faster growing individuals may be more susceptible to single-species "outbreaks", and therefore increasingly likely to become dominance-orientated. Whatever the reason, these two significant trends call into question the wisdom of simply using the mean and 95% confidence limits of the low effort treatment rectangles as reference levels for these two metrics. EcoQOs set against such reference levels for these metrics may simply be unattainable, given changes in the factor(s) responsible for the significant trends in areas where fishing effort has been so low that fishing is unlikely to have been the cause. In such circumstances, the regression relationship for the low effort treatment might instead be used to set reference levels more appropriate for the present.
For all 11 metrics, the temporal trend lines for the medium and high effort treatments crossed the mean and lower 95% confidence limit values for the low effort treatment, suggesting that the point in time when the community started to deviate away from an undisturbed state was contained within the timespan of the data time-series. However, while these deviation points occurred relatively early on in the time-series with respect to the seven size-based and life-history-based metrics, this was not the case with the five metrics of species diversity. Indeed, for all five species diversity metrics, early data point values for both the medium and high effort treatments were considerably greater than the upper 95% confidence limit value for the low effort treatment. It would seem that species richness and diversity in the medium and high effort ICES statistical rectangles have not been perturbed away from levels equivalent to the low effort treatment means, but instead may have deviated away from values initially much higher than these. Therefore, the low effort treatment mean values of these five metrics do not represent the non-disturbed state. If this is the case for those metrics, then it also begs the question of whether had the data set started even earlier, would the same have been observed for the size- and life-history-based metrics? Even if there is some doubt as to whether the low effort treatment data actually represent true reference levels within the strict definition of the term, they still provide a clear basis for setting reference directions, sensu Jennings and Dulvy (2005).
Finally, it must be recognized that groundfish survey data provide a gear-biased view of the fish communities they sample. The "pictures" obtained are views of the real world seen through the "lens" of the sampling gear used. Therefore, trends in the metrics presented here are the trends detected using an Aberdeen 48-ft demersal trawl. This does not invalidate the conclusions drawn: the signals are real. However, it is important to apply these community metrics to a variety of different surveys in order to determine how universal the signals are. Different groundfish surveys would in all likelihood pick up the same trends (Piet and Jennings, 2005), but the actual metric values may very well differ between different data sets derived from different trawl gears. This becomes an important issue when it comes to setting EcoQOs. The data set analysed here is perhaps the longest lasting continuous groundfish survey that has been carried out in the North Sea, long-lasting enough apparently to have detected the deviation point away from the non-anthropogenically affected state for many of the metrics examined. Therefore, it may indeed be possible to use this data set to identify reference levels against which EcoQOs might be set. However, this survey was discontinued in 1997. It is therefore no longer available as a tool to monitor progress towards such EcoQOs. Comparative studies between different trawl surveys are essential as progress towards EcoQOs set on the basis of the SAGFS will have to be monitored using data gathered using different sampling fishing gears.
|
Acknowledgements |
|---|
We thank colleagues at two ICES Working Groups, Ecosystem Effects of Fishing Activities (WGECO) and Fish Ecology (WGFE), for their contribution to the development of the ideas expressed here, specifically Chris Frid, Jake Rice, Nils Daan, and Gerjan Piet. The work was carried out under Service Level agreement MF07A to SEERAD. It constitutes part contributions to two CEFAS projects "Using Historical Data Sets to Investigate Impacts of Fishing and Climate Change on Demersal Fish Assemblages" (MF0730) and "Development and Testing of Ecological Indicators and Models to Monitor and Predict Ecosystem Effects of Fishing" (MF0731). It also represents a contribution to the EC INDECO project "Developing Indicators of Environmental Performance of the Common Fisheries Policy" (FP6-2003-SSP-3-513754). We are most grateful to Simon Jennings who was closely involved in the evolution of this paper, not only through his input at the two ICES working groups and through his role as principal investigator in CEFAS project MF0731, but also as a long-term collaborator in studies of this nature. Finally, we thank two anonymous referees and editor Verena Trenkel for their helpful comments on earlier drafts of the manuscript.
|
References |
|---|
|
TopIntroductionMethodsResultsDiscussion and conclusionsReferences |
|---|
-
Badalamenti G., Anna G.D., Pinnegar J.K., Polunin N.V.C. (2002) Size-related trophodynamic changes in three target fish species recovering from intensive trawling. Marine Biology 141:561–570.[CrossRef]
Beddington J.R. (1995) The primary requirements. Nature 374:213–214.[CrossRef]
Bergen Declaration. (2002) Ministerial Declaration, Fifth International Conference on the Protection of the North SeaBergen, Norway.
Beverton R.J.H. and Holt S.J. (1957) On the dynamics of exploited fish populations. Fishery Investigations Series II 19:533.
Bianchi G., Gislason H., Graham K., Hill L., Jin X., Koranteng K., Manickchand-Heileman S., Payá I., Sainsbury K., Sanchez F., Zwanenburg K. (2000) Impact of fishing on size composition and diversity of demersal fish communities. ICES Journal of Marine Science 57:558–571.
Blanchard F., LeLoc'k F., Hily C., Boucher J. (2004) Fishing effects on diversity, size and community structure of the benthic invertebrate and fish megafauna on the Bay of Biscay coast of France. Marine Ecology Progress Series 280:249–260.[Web of Science]
Blanchard J.L., Dulvy N.K., Jennings S., Ellis J.R., Pinnegar J.K., Tidd A., Kell L.T. (2005) Do climate and fishing influence size-based indicators of Celtic Sea fish community structure. ICES Journal of Marine Science 62:405–411.
Buckland S.T., Magurran A.E., Green R.E., Fewster R.M. (2005) Monitoring changes in biodiversity through composite indices. Philosophical Transactions of the Royal Society B 360:243–254.[CrossRef]
Bundy A., Fanning P., Zwanenburg C.T. (2005) Balancing exploitation and conservation of the eastern Scotian Shelf ecosystem: application of a 4D ecosystem exploitation index. ICES Journal of Marine Science 62:503–510.
Carpenter S.R., Kitchell J.F., Hodgson J.R., Cochran P.A., Elser J.J., Lodge D.M., Kretchmer D., He X., von Ende C.N. (1987) Regulation of lake primary productivity by food web structure. Ecology 68:1863–1876.[CrossRef][Web of Science]
Chadwick J.W. and Canton S.P. (1984) Inadequacy of diversity indices in discerning metal mine drainage effects on a stream invertebrate community. Water, Air and Soil Pollution 22:217–223.
Clark R.A. and Frid C.L.J. (2001) Long-term changes in the North Sea ecosystem. Environmental Reviews 9:131–187.[CrossRef]
Clarke K.R. and Warwick K.M. (2001) Change in Marine Communities: an Approach to Statistical Analysis and Interpretation 2nd edn (PRIMER-E, Plymouth, UK).
Colwell R.K., Mao C.X., Chank J. (2004) Interpolating, extrapolating, and comparing incidence-based species accumulation curves. Ecology 85:2717–2727.[CrossRef][Web of Science]
Connell J.H. (1975) Some mechanisms producing structure in natural communities: a model and evidence from field experiments. In Cody M.L. and Diamond J.M. (Eds.). Ecology and Evolution of Communities(The Belknap Press of Harvard University, Cambridge, Massachusetts) pp. 460–491.
Connell J.H. (1978) Diversity in tropical rain forests and coral reefs. Science 199:1302–1310.
Conover D.O. and Munch S.B. (2002) Sustaining fisheries yields over evolutionary time scales. Science 297:94–96.
Cury P., Shannon L.J., Shin Y-J. (2003) The functioning of marine ecosystems: a fisheries perspective. In Sinclair M. and Valdimarsson G. (Eds.). Responsible Fisheries in the Marine Ecosystem(CAB International, Wallingford) pp. 103–123.
Cury P.M. and Christensen V. (2005) Quantitative ecosystem indicators for fisheries management: introduction. ICES Journal of Marine Science 62:307–310.
Cury P.M., Shannon L.J., Roux J-P., Daskalov G.M., Jarre A., Moloney C.L., Pauly D. (2005) Trophodynamic indicators for an ecosystem approach to fisheries. ICES Journal of Marine Science 62:430–442.
Daan N., Bromley P.J., Hislop J.R.G., Nielsen N.A. (1990) Ecology of North Sea fish. Netherlands Journal of Sea Research 26:343–386.
Daan N., Gislason H., Pope J.G., Rice J.C. (2005) Changes in the North Sea fish community: evidence of indirect effects of fishing. ICES Journal of Marine Science 62:177–188.
Emmerson M. and Huxham M. (2002) How can marine ecology contribute to the biodiversity–ecosystem functioning debate? In Loreau M., Naeem S., Inchausti P. (Eds.). Biodiversity and Ecosystem Functioning: Synthesis and Perspectives(Oxford University Press, Oxford, UK) pp. 139–146.
Fulton E.A., Smith A.D.M., Punt A.E. (2005) Which ecological indicators can robustly detect effects of fishing. ICES Journal of Marine Science 62:540–551.
Garcia S.M. and Cochrane K.L. (2005) Ecosystem approach to fisheries: a review of implementation guidelines. ICES Journal of Marine Science 62:311–318.
Gascuel D., Bozec Y-M., Chassot E., Colomb A., Laurans M. (2005) The trophic spectrum: theory and application as an ecosystem indicator. ICES Journal of Marine Science 62:443–452.
Gislason H., Sinclair M., Sainsbury K., O'Boyle R. (2000) Symposium overview: incorporating ecosystem objectives within fishery management. ICES Journal of Marine Science 57:468–475.
Greenstreet S.P.R. and Hall S.J. (1996) Fishing and the ground-fish assemblage structure in the north-western North Sea: an analysis of long-term and spatial trends. Journal of Animal Ecology 65:577–598.[CrossRef][Web of Science]
Greenstreet S.P.R. and Rogers S.I. (2000) Effects of fishing on non-target fish species. In Kaiser M.J. and de Groot B. (Eds.). Effects of Fishing on Non-Target Species and Habitats: Biological, Conservation and Socio-economic Issues(Blackwell Science, Oxford, UK) pp. 217–234.
Greenstreet S.P.R., Spence F.E., McMillan J.A. (1999) Fishing effects in northeast Atlantic shelf seas: patterns in fishing effort, diversity and community structure. 5. Changes in structure of the North Sea groundfish assemblage between 1925 and 1996. Fisheries Research 40:153–183.[CrossRef][Web of Science]
Greenstreet S.P.R., Spence F.E., Shanks A.M., McMillan J.A. (1999) Fishing effects in northeast Atlantic shelf seas: patterns in fishing effort, diversity and community structure. 2. Trends in fishing effort in the North Sea by U.K. registered vessels landing in Scotland. Fisheries Research 40:107–124.[CrossRef][Web of Science]
Greenstreet S. P. R., Shanks A. M., Buckett B-E. (2006) Trends in fishing activity in the North Sea by U.K. registered vessels landing in Scotland over the period 1960 to 1998. Fisheries Research Services Collaborative Reports, 02/06. 27 pp.
Grift R.E., Rijnsdorp A.D., Barot S., Heino M., Dieckmann U. (2003) Fisheries-induced trends in reaction norms for maturation in North Sea plaice. Marine Ecology Progress Series 257:247–257.[Web of Science]
Hall S.J. and Mainprize B. (2004) Towards ecosystem-based fisheries management. Fish and Fisheries 5:1–20.[CrossRef][Web of Science]
Heath M.R. (2005) Changes in the structure and function of the North Sea fish foodweb, 1973–2000, and the impacts of fishing and climate. ICES Journal of Marine Science 62:847–868.
Hill M.O. (1973) Diversity and evenness: a unifying notation and its consequences. Ecology 54:427–432.[CrossRef][Web of Science]
Hurlbert S.H. (1971) The nonconcept of species diversity: a critique and alternative parameters. Ecology 52:577–586.[CrossRef][Web of Science]
Huston M.A. (1994) Biological Diversity: The Coexistence of Species on Changing Landscapes(Cambridge University Press, Cambridge, UK) 681 pp.
Huston M.A. and McBride A.C. (2002) Evaluating the relative strengths of biotic versus abiotic controls on ecosystem processes. In Loreau M., Naeem S., Inchausti P. (Eds.). Biodiversity and Ecosystem Functioning: Synthesis and Perspectives(Oxford University Press, Oxford, UK) pp. 47–60.
ICES. (2001) Report of the ICES Advisory Committee on Ecosystems. ICES Cooperative Research Report, 249. 75 pp.
ICES. (2004) Report of the Working Group on the assessment of Demersal Stocks in the North Sea and Skagerrak. ICES Document CM 2004 ACFM: 07. 555 pp.
Jennings S., Alvsv
g J., Cotter A.J., Ehrich S., Greenstreet S.P.R., Jarre-Teichmann A., Mergardt N., Rijnsdorp A.D., Smedstad O. (1999) Fishing effects in northeast Atlantic shelf seas: patterns in fishing effort, diversity and community structure. 3. International fishing effort in the North Sea: an analysis of spatial and temporal trends. Fisheries Research 40:125–134.[CrossRef][Web of Science]
Jennings S. and Blanchard J.L. (2004) Fish abundance with no fishing: predictions based on macro-ecological theory. Journal of Animal Ecology 73:632–642.[CrossRef][Web of Science]
Jennings S. and Dulvy N.K. (2005) Reference points and reference directions for size-based indicators of community structure. ICES Journal of Marine Science 62:397–404.
Jennings S., Greenstreet S.P.R., Hill L., Piet G.J., Pinnegar J., Warr K.J. (2002) Long-term trends in the trophic structure of the North Sea fish community: evidence from stable isotope analysis, size-spectra and community metrics. Marine Biology 141:1085–1097.[CrossRef]
Jennings S., Greenstreet S.P.R., Reynolds J. (1999) Structural change in an exploited fish community: a consequence of differential fishing effects on species with contrasting life histories. Journal of Animal Ecology 68:617–627.[CrossRef][Web of Science]
Jennings S., Pinnegar J.K., Polunin N.V.C., Boon T. (2001) Weak cross species relationships between body size and trophic level belie powerful size-based trophic structuring in fish communities. Journal of Animal Ecology 70:934–944.[CrossRef][Web of Science]
Jennings S., Pinnegar J.K., Polunin N.V.C., Warr K.J. (2002) Linking size-based and trophic analyses of benthic community structure. Marine Ecology Progress Series 226:77–85.[Web of Science]
Jennings S. and Reynolds J.D. (2000) Impacts of fishing on diversity: from pattern to process. In Kaiser M.J. and de Groot B. (Eds.). Effects of Fishing on Non-Target Species and Habitats: Biological, Conservation and Socio-economic Issues(Blackwell Science, Oxford, UK) pp. 235–250.
Jennings S., Reynolds J.D., Mills S.C. (1998) Life history correlates of responses to fisheries exploitation. Proceedings of the Royal Society of London 265:1–7.[Medline]
Jennings S. and Warr K.J. (2003) Smaller predator–prey body size ratios in longer food chains. Proceedings of the Royal Society of London, Series B 270:1413–1417.[Medline]
Jennings S., Warr K.J., Greenstreet S.P.R., Cotter A.J.R. (2000) Spatial and temporal patterns in North Sea fishing effort. In Kaiser M.J. and de Groot B. (Eds.). Effects of Fishing on Non-Target Species and Habitats: Biological, Conservation and Socio-economic Issues(Blackwell Science, Oxford, UK) pp. 3–14.
Jennings S., Warr K.J., Mackinson S. (2002) Use of size-based production and stable isotope analyses to predict trophic transfer efficiencies and predator–prey body mass ratios in food webs. Marine Ecology Progress Series 240:11–20.[Web of Science]
Kerr S.R. and Dickie L.M. (2001) The Biomass Spectrum: A Predator Prey Theory of Aquatic Production(Columbia University Press, New York) 320 pp.
Lanters R. L. P., Skjoldal H. R., Noji T. T. (1999) Ecological Quality Objectives for the North Sea, 1–3 September 1999. Scheveningen, Netherlands. Joint Report, 99, Netherlands Ministry of Transport, Public Works and Water Management, National Institute for Coastal and Marine Management (RIKZ)/Norwegian Institute of Marine Research.
Law R. and Grey D.R. (1989) Evolution of yields from populations with age-specific cropping. Evolutionary Ecology 3:343–359.[CrossRef][Web of Science]
Lekve K., Ottersen G., Stenseth N.C., Gjosaeter J. (2002) Length dynamics in juvenile coastal Skagerrak cod: effects of biotic and abiotic factors. Ecology 86:1676–1688.
Loreau M., Downing A., Emmerson M., Gonzalez A., Huches J., Inchausti P., Joshi J., Norberg J., Sala O. (2002) A new look at the relationship between diversity and stability. In Loreau M., Naeem S., Inchausti P. (Eds.). Biodiversity and Ecosystem Functioning: Synthesis and Perspectives(Oxford University Press, Oxford, UK) pp. 79–91.
MacArthur R.H. and Wilson E.O. (1967) The Theory of Island Biogeography(Princeton University Press, Princeton) 203 pp.
Magurran A.E. (1988) Ecological Diversity and its Measurement(Chapman and Hall, London) 179 pp.
Minagawa M. and Wada E. (1984) Stepwise enrichment of 15N along food chains: further evidence and the relation between 15N and animal age. Geochimica et Cosmochimica Acta 48:1135–1140.[CrossRef][Web of Science]
Mooney H.A. (2002) The debate on the role of biodiversity in ecosystem functioning. In Loreau M., Naeem S., Inchausti P. (Eds.). Biodiversity and Ecosystem Functioning: Synthesis and Perspectives(Oxford University Press, Oxford, UK) pp. 12–17.
Murawski S.A. and Idoine J.S. (1992) Multispecies size composition: a conservative property of exploited fishery systems. Journal of Northwest Atlantic Fisheries Science 14:79–85.
Olsen E.M., Lilly G.R., Heino M., Morgan M.J., Brattey J., Dieckmann U. (2005) Assessing changes in age and size at maturation in collapsing populations of Atlantic cod (Gadus morhua). Canadian Journal of Fisheries and Aquatic Sciences 62:811–823.
Ottersen G. and Loeng H. (2000) Covariability in early growth and year-class strength of Barents Sea cod, haddock, and herring: the environmental link. ICES Journal of Marine Science 57:339–348.
Paine R.T. (1974) Intertidal community structure. Experimental studies on the relationship between a dominant competitor and its principal predator. Oecologia 15:93–120.[CrossRef][Web of Science]
Pauly D. and Christensen V. (1995) Primary production required to sustain global fisheries. Nature 374:255–257.[CrossRef]
Pauly D., Christensen V., Dalsgaard J., Froese R., Torres F. (1998) Fishing down marine food webs. Science 279:860–863.
Pauly D. and Maclean J. (2003) In a Perfect Ocean: the State of Fisheries and Ecosystems in the North Atlantic Ocean(Island Press, Washington DC) 175 pp.
Pauly D., Palomares M.L., Froese R., Pascualita S., Vakily M., Preikshot D. (2001) Fishing down Canadian aquatic food webs. Canadian Journal of Fisheries and Aquatic Sciences 58:51–62.
Petchey O.L., Morin P.J., Hulot F.D., Loreau M., McGrady-Steed J., Naeem S. (2002) Contributions of aquatic model systems to our understanding of biodiversity and ecosystem functioning. In Loreau M., Naeem S., Inchausti P. (Eds.). Biodiversity and Ecosystem Functioning: Synthesis and Perspectives(Oxford University Press, Oxford, UK) pp. 127–138.
Pielou E.C. (1975) Ecological Diversity(Wiley, New York) 165 pp.
Pielou E.C. (1969) An Introduction to Mathematical Ecology(Wiley, New York) 286 pp.
Piet G.J. and Jennings S. (2005) Response of potential fish community indicators to fishing. ICES Journal of Marine Science 62:214–225.
Pinnegar J.K., Jennings S., O'Brien C.M., Polunin N.V.C. (2002) Long-term changes in trophic level of the Celtic Sea fish community and fish market price distribution. Journal of Applied Ecology 39:377–390.[CrossRef][Web of Science]
Pope J.G. (1991) The ICES Multispecies Assessment Working Group: evolution, insights and future problems. In Daan N. and Sissenwine M.P. (Eds.). Models Relevant to Management of Living Resources 193: pp. 22–33 ICES Marine Science Symposia.
Raffaelli D., van der Putten W.H., Persson L., Wardle D.A., Petchey O.L., Koricheva J., van der Heijden M., Mikola J., Kennedy T. (2002) Multi-trophic dynamics and ecosystem processes. In Loreau M., Naeem S., Inchausti P. (Eds.). Biodiversity and Ecosystem Functioning: Synthesis and Perspectives(Oxford University Press, Oxford, UK) pp. 139–146.
Reid P.C., Lancelot C., Gieskes W.W.C., Hagmeier E., Weichart G. (1990) Phytoplankton of the North Sea and its dynamics: a review. Netherlands Journal of Sea Research 26:295–331.
Rice J. (2003) Environmental health indicators. Ocean and Coastal Management 46:235–259.[CrossRef]
Rice J. and Gislason H. (1996) Patterns of change in the size spectra of numbers and diversity of the North Sea fish assemblage, as reflected in surveys and models. ICES Journal of Marine Science 53:1214–1225.
Rice J.C. and Rochet M-J. (2005) A framework for selecting a suite of indicators for fisheries management. ICES Journal of Marine Science 62:516–527.
Ricker W.E. (1995) Trends in the average size of Pacific salmon in Canadian catches. In Beamish R.J. (Ed.). Climate Change and Northern Fish Populations pp. 593–602.
Rijnsdorp A. (1993) Fisheries as a large-scale experiment on life-history evolution: disentangling phenotypic and genetic effects in changes in maturation and reproduction of North Sea plaice, Pleuronectes platessa L. Oecologia 96:391–401.[CrossRef][Web of Science]
Robinson J.V. and Sandgren C.D. (1984) An experimental evaluation of diversity indices as environmental discriminators. Hydrobiologia 108:187–196.[CrossRef][Web of Science]
Rochet M-J. and Trenkel V.M. (2003) Which community indicators can measure the impact of fishing? A review and proposals. Canadian Journal of Fisheries and Aquatic Sciences 60:86–99.
Rogers S.I., Clarke K.R., Reynolds J.D. (1999) The taxonomic distinctness of coastal bottom-dwelling fish communities of the Northeast Atlantic. Journal of Animal Ecology 68:769–782.[CrossRef][Web of Science]
Rogers S.I. and Ellis J.R. (2000) Changes in the demersal fish assemblages of British coastal waters during the 20th century. ICES Journal of Marine Science 57:866–881.
Rogers S.I., Maxwell D., Rijnsdorp A.D., Damm U., Vanhee W. (1999) Fishing effects in northeast Atlantic shelf seas: patterns in fishing effort, diversity and community structure. 4. Can comparisons of species diversity be used to assess human impacts on demersal fish faunas? Fisheries Research 40:135–152.[CrossRef][Web of Science]
Rogers S.I., Rijnsdorp A.D., Damm U., Vanhee W. (1998) Demersal fish populations in the coastal waters of the UK and continental NW Europe from beam trawl survey data collected from 1990–1995. Journal of Sea Research 39:79–102.[CrossRef]
Rosenzweig M.L. (1995) Species Diversity in Space and Time(Cambridge University Press, Cambridge) 436 pp.
Sainsbury K. and Sumaila U.R. (2001) Incorporating ecosystem objectives into management of sustainable marine fisheries including "best practice" reference points and use of marine protected areas. In Sinclair M. (Ed.). Responsible Fisheries in the Marine Ecosystem(CABI Publishing, Oxford) pp. 343–461.
Shin Y-J., Rochet M-J., Jennings S., Field J.G., Gislason H. (2005) Using size-based indicators to evaluate the ecosystem effects of fishing. ICES Journal of Marine Science 62:348–396.
Soetaert K. and Heip C. (1990) Sample-size dependence of diversity indices and the determination of sufficient sample size in a high-diversity deep-sea environment. Marine Ecology Progress Series 59:305–307.[Web of Science]
Southwood T.R.E. (1978) Ecological Methods: With Particular Reference to the Study of Insect Populations(Chapman and Hall, London).
Sparholt H. (1990) An estimate of the total biomass of fish in the North Sea. Journal du Conseil International pour l'Exploration de la Mer 46:200–210.
Stobberup K.A., Inejih C.A.O., Traoré S., Monteiro C., Amorim P., Erzini K. (2005) Analysis of size-spectra off northwest Africa: a useful indicator in tropical areas? ICES Journal of Marine Science 62:424–429.
Stransky C. (1998) A comparison of biomass estimates obtained from small and large scale groundfish surveys in the North Sea. MSc thesis, University of Aberdeen, UK.
Tilman D. and Downing J.A. (1994) Biodiversity and stability in grasslands. Nature 367:363–365.[CrossRef]
Tilman D., Knops J., Wedin D., Reich P. (2002) Experimental and observational studies of diversity, productivity and stability. In Kinzig A.P.S. and Tilman D. (Eds.). Functional Consequences of Biodiversity: Experimental Progress and Theoretical Extensions(Princeton University Press, Princeton, NJ) pp. 42–70.
Tilman D., Knops J., Wedin D., Reich P. (2002) Plant diversity and composition: effects on productivity and nutrient dynamics of experimental grasslands. In Loreau M., Naeem S., Inchausti P. (Eds.). Biodiversity and Ecosystem Functioning: Synthesis and Perspectives(Oxford University Press, Oxford, UK) pp. 21–35.
Trenkel V.M. and Rochet M-J. (2003) Performance indicators derived from abundance estimates for detecting the impact of fishing on a fish community. Canadian Journal of Fisheries and Aquatic Sciences 60:67–85.
Velleman P.F. and Welsch R.E. (1981) Efficient computing of regression diagnostics. The American Statistician 35:234–242.[CrossRef]
Verity P.G. and Smetacek P. (1996) Organism life cycles, predation and the structure of marine pelagic ecosystems. Marine Ecology Progress Series 130:277–293.[Web of Science]
Washington H.G. (1984) Diversity, biotic and similarity indices: a review with special relevance to aquatic ecosystems. Water Research 18:653–694.
Wilderbuer T.K., Hollowed A.B., Ingraham W.J., Spencer P.D., Connors M.E., Bond N.A., Walters G.E. (2002) Flatfish recruitment response to decadal climatic variability and ocean conditions in the eastern Bering Sea. Progress in Oceanography 55:235–247.[CrossRef][Web of Science]
Wilsey B.J., Chalcraft D.R., Bowles C.M., Willig M.R. (2005) Relationships among indices suggest that richness is an incomplete surrogate for grassland biodiversity. Ecology 86:1178–1184.[Web of Science]
Yang J. (1982) An estimate of the fish biomass in the North Sea. Journal du Conseil International pour l'Exploration de la Mer 40:161–172.
Zwanenburg K.C.T. (2000) The effects of fishing on demersal fish communities of the Scotian Shelf. ICES Journal of Marine Science 57:503–509.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. J. Piet, R. van Hal, and S. P. R. Greenstreet Modelling the direct impact of bottom trawling on the North Sea fish community to derive estimates of fishing mortality for non-target fish species ICES J. Mar. Sci., June 14, 2009; (2009) fsp162v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. P. R. Greenstreet, G. J. Holland, T. W. K. Fraser, and V. J. Allen Modelling demersal fishing effort based on landings and days absence from port, to generate indicators of "activity" ICES J. Mar. Sci., June 1, 2009; 66(5): 886 - 901. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Rice A generalization of the three-stage model for advice using the precautionary approach in fisheries, to apply broadly to ecosystem properties and pressures ICES J. Mar. Sci., April 1, 2009; 66(3): 433 - 444. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. M. Fraser, S. P. R. Greenstreet, and G. J. Piet Selecting MPAs to conserve groundfish biodiversity: the consequences of failing to account for catchability in survey trawls ICES J. Mar. Sci., January 1, 2009; 66(1): 82 - 89. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. P. R. Greenstreet, H. M. Fraser, and G. J. Piet Using MPAs to address regional-scale ecological objectives in the North Sea: modelling the effects of fishing effort displacement ICES J. Mar. Sci., January 1, 2009; 66(1): 90 - 100. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. P. R. Greenstreet Biodiversity of North Sea fish: why do the politicians care but marine scientists appear oblivious to this issue? ICES J. Mar. Sci., November 1, 2008; 65(8): 1515 - 1519. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Barry and H. L. Rees Use of simulated data as a tool for testing the performance of diversity indices in response to an organic enrichment event ICES J. Mar. Sci., November 1, 2008; 65(8): 1456 - 1461. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Heath Comment on "A Global Map of Human Impact on Marine Ecosystems" Science, September 12, 2008; 321(5895): 1446b - 1446b. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Gislason, J. G. Pope, J. C. Rice, and N. Daan Coexistence in North Sea fish communities: implications for growth and natural mortality ICES J. Mar. Sci., May 1, 2008; 65(4): 514 - 530. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. M. Fraser, S. P. R. Greenstreet, R. J. Fryer, and G. J. Piet Mapping spatial variation in demersal fish species diversity and composition in the North Sea: accounting for species- and size-related catchability in survey trawls ICES J. Mar. Sci., May 1, 2008; 65(4): 531 - 538. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. M. Fraser, S. P. R. Greenstreet, and G. J. Piet Taking account of catchability in groundfish survey trawls: implications for estimating demersal fish biomass ICES J. Mar. Sci., December 1, 2007; 64(9): 1800 - 1819. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||