Page History

Indicator summary

Summary of indicator structure and function

Examples of how the indicators is used for ecosystem management and ecosystem status and trends

Definition and/or background

The following is from Fulton et al 2004a -

Mortality rates can be determined from trawl catches for species and or aggregates of species using a range of ‘rapid assessment’ methods. Once determined, mortality rates can be compared to suggested reference points (Die and Caddy 1997) or modelled against increasing fishing effort to determine reference points of vulnerability for species (q-jeopardy) susceptible to population decline from trawling (Pope et al. 2000).

Die and Caddy (1997) suggest that total mortality Z (calculated using the expression of Beverton and Holt 1957 in Die and Caddy 1997) could be used to monitor a fishery’s development. Z calculated in this way is given by:

where L_∞ and Kare the von Bertalanffy growth parameters, is the average length of the catch and L_c is the size of first capture. This formulation is particularly useful as it does not require estimation of natural mortality (M), which can be difficult to determine. This value of Z can then be compared to standard fisheries limit reference points such as Z_0.1 (total mortality), Z_mbp(total mortality leading to maximum biological production) and Z*, which is an upper limit for a sustainable Z. Z* assumes <L_m (where L_m is the size of 50% maturity) and is given by:

Using Z calculated in this way builds on existing fisheries knowledge, while avoiding many of the sampling and uncertainty issues that plague some stock indices. The use of Z as an indicator may be particularly useful at a population level, it was found to perform reasonably well when applied to the Celtic sea by Trenkel and Rochet (2003). It must be remembered however, that it does not indicate the ecological situation (relative predation, food availability and competitive pressure) a species will be under once its stocks have shifted from unfished levels (Die and Caddy 1997). Whereas the methods of Die and Caddy (1997) concentrate primarily on harvested groups, the methods proposed by Pope et al. (2000) estimate fishing mortality rates for non-target species. There is often little data for non-target species, which can make it hard to determine the impact of fishing on them. Pope et al. 2000 show that via the inversion of the fishing mortality rate it is possible to estimate the impacts of fishing related mortality on populations (this is known as the potential jeopardy level). F’(φ) is an estimate of the fishing mortality that would reduce the spawning stock to φ% of its unfished levels and is also known as the φ% potential jeopardy level of fishing mortality. The potential jeopardy level could be used as a reference point, but is also a means of assessing the impacts of current levels of mortality on populations (and thus the system).

Extended length cohort analysis with potential jeopardy level

This methodology factors in mortality for each sex and the life-history parameters of age at first maturity for females. This is important because growth rates differ between sexes and this makes it necessary to have sex-based information as mortality has an impact on spawning-stock biomass (SSB).  Since it would be costly and time consuming to sex each non-target fish, and since many existing data sets don’t have that information, a model has been developed based on the assumption of an initial population size being the same for each sex and sustaining an arbitrary level of mortality for each length group through time. The population surviving at each length and the catch by length group are then calculated and an objective function based on these is minimised to find the initial population size and mortality per length group that satisfies the data and the assumptions. For example, the mortality of the catch for each species taken in the North Sea was estimated using the following general relationship between natural mortality rate (M) and weight (W), which was calculated per species from ICES 1988 data in (Pope et al. 2000):

...

Weighted swept area method

This is a rapid assessment method that can be applied to a fishery to generate an indicator of non-target species mortality when limited data are available. It assumes that the local mortality (F) is proportional to the fraction of the area swept by fishing gear. The area fished is calculated as a proportion of the total area (e.g. percentage of an ICES fishery rectangle) and by assuming fishing is a random with respect to non-target species and that catchability is 1 (all fish in a track are caught), the annual local fishing rate (F_r), and total mortality (F_tot) can be calculated. To explore its effectiveness, results produced by this method when using the same data sets for dab and grey gurnard used to explore the extended length cohort analysis with potential jeopardy level (above), were compared with the results of the extended length cohort analysis. It was found that the results were similar for both methods and that the methods gave realistic estimates of the minimum size of capture and levels of fishing mortality (Pope et al. 2000). The advantage to the swept area method is that it does not require sampling of commercial by-catches and the required data on the distribution and abundance of non-target species are often available from other sources (Pope et al. 2000).

The simpler form of the assumptions used in this method mean that the jeopardy level can be expressed in terms of simple life-history parameters. In turn this allows for the calculation of the predicted response surface linking potential jeopardy level with some of the life-history parameters (Figure 5.10). Moreover, by using dimensionless rations (e.g. M/K) the vulnerability of species can be based on growth rather than life-history parameters, which are much harder to measure. The existence of simple relationships, like the one in Figure 1, may make it possible to give estimates of vulnerability for even the most poorly known species.

Figure 1: Potential jeopardy level of fishing mortality increases with K and the M/K ratio. Species with ‘faster’ life histories are less vulnerable to fishing mortality (Pope et al. 2000). With M = natural mortality rate at length; K = growth parameter by sex; F = fishing mortality rate at length. (Figure provided by Fulton et al 2004a).

Attribute

Community structure, Population structure

Purpose

Fisheries

Data required

The following is from Fulton et al 2004a -

Total mortality (Z):

• Natural (M) and fishing (F) mortality for species of interest
• Alternatively, mean length of catch, length at first capture and von Bertalanffy parameters K (growth) and L_∞ (mean maximum length)_.

Extended length cohort analysis:

• Estimate of the steady-state (or average) catch-at-length distribution for each species
• Growth and female maturity parameters for each species

Weighted swept area method:

• Catch rates by species and area
• Commercial effort data (preferably by fleet)

Ecosystem applicability

The following is from Fulton et al 2004a -

Both measures are potentially applicable to all ecosystems.

Robustness

The following is from Fulton et al 2004a -

Total mortality: Potentially high, though it can be impacted by factors other than fishing. If good data on life history parameters (especially natural mortality) are available then total mortality should be a robust indicator, as it is easily understood and potential reference points have already been identified.

Extended length cohort analysis: Potentially high, especially if good data on growth and life-history parameters are available. The values given by this method in the analysis of Pope et al. (2000) seem plausible and are realistic, but like all new methods it should be tested stringently before coming into general use (Pope et al. 2000).

Weighted swept area method: Medium, as the upper estimates of mortality for fully selected size classes are generated, but the methodology does not provide information on which sizes are fully selected. While the selection characteristics may be inferred from other species, invertebrates will probably require more study. Moreover, the assumptions underlying this methodology are much weaker than for other methods (particularly the assumption of random tows) and this may make the method inappropriate for at least some species. However, with regard to trawl fishing, the index is relatively reliable for fish species adapted to life on the sea-bed as the assumptions (particularly the assumption that catchability is one) are fairly realistic for these groups (Pope et al. 2000).

Current status and trends

The following is from Fulton et al 2004a -

Extended length cohort analysis with potential jeopardy level

The model was tested on discards of dab (Limanda limanda) and Grey gurnard (Eutrigula gurnadus) from the Dutch beam trawl fleet data 1977-1983. Results provided an index of fishing mortality rate by length - there was no mortality below 12cm, mortality rose quickly between lengths of 15-25cm (Figure 2). Estimates of multipliers for North Sea fisheries that would generate F’(20%) were also calculated, and were found to be 279 for dab and 10 for Grey gurnard. Together these findings indicate that these fish populations should not be vulnerable at current levels of bycatch. This is because the females of both species first mature between 12-19cm (before they are fully recruited to the fishery) and so it was predicted that spawning potential (SSB) was not badly eroded and that the fish species are not vulnerable to over-fishing under current catch rates and that it would take extreme increase in F to reduce these fish species to the F’(20%) levels.

Figure 2: Fishing mortality rate by length group for North Sea stocks of Dab (solid line) and Grey Gurnard (dashed line) calculated using the differential growth version of extended length cohort analysis (after Pope et al. 2000).  (Figure provided by Fulton et al 2004a).

References

Fulton, E.A., Smith, A.D.M., Webb, H., and Slater, J. (2004a) Ecological indicators for the impacts of fishing on non-target species, communities and ecosystems: Review of potential indicators. AFMA Final Research Report, report Number R99/1546.

References that Fulton et al uses for this indicator:

Die, D. J., and J. F. Caddy. 1997. Sustainable yield indicators from biomass: are there appropriate reference points for use in tropical fisheries? Fisheries Research 32: pp 69-79.

Pope, J.G., D. S. MacDonald, N. Daan, J. D. Reynolds, and S. Jennings. 2000. Gauging the impact of fishing mortality on non-target species. ICES Journal of Marine Science 57: pp 689-96.

Trenkel, V.M., and M.-J. Rochet. 2003. Performance of indicators derived from abundance estimates for detecting the impact of fishing on a fish community. Canadian Journal of Fisheries and Aquatic Sciences 60: pp 67-85.

Background reading

Fulton, E.A., Fuller,M., Smith, A.D.M., and Punt, A. (2004) Ecological indicators of the ecosystem effects of fishing: Final report. AFMA Final Research Report, report Number R99/1546.

Citation

Page created by:

Last modified on:

Versions: