SEARCH

SEARCH BY CITATION

Keywords:

  • CITES;
  • extinction risk;
  • marine fisheries;
  • minimum viable population;
  • Red List

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    Threat listing of exploited marine species has been controversial because of the scientific uncertainty of extinction risk as well as the social, economic and political costs of management procedures that may be triggered by designation of species as threatened.
  • 2
    We applied three sets of threat criteria to 76 stocks (populations) of 21 exploited marine fish and invertebrate species. Two criteria sets were based on decline rates: World Conservation Union (IUCN A1) and the American Fisheries Society (AFS). The third set of criteria, based on population viability (IUCN E), was assessed using non-parametric simulation and two diffusion approximation methods.
  • 3
    We compared extinction risk outcomes (threatened or not) against the exploitation status of each stock as reported in fish stock assessments (inside or outside safe biological limits). For each combination of threat and exploitation we assessed the rate of hits, misses and false alarms.
  • 4
    Our analyses suggest that decline rate criteria provide risk categorizations consistent with population viability analyses when applied to exploited marine stocks. Nearly a quarter of the fish and invertebrate populations (n = 18) considered met one or more of the threat criteria.
  • 5
    None of the threat metrics produced false alarms, where sustainably exploited stocks were categorized as threatened. The quantitative IUCN E metrics produced higher hit rates than the decline rate metrics (IUCN A1 and AFS) and all of the metrics produced similar miss rates. However, the IUCN E methods could be applied to fewer stocks (12–14) compared with IUCN A1 decline rate and AFS criteria, both of which could be applied all 76 stocks.
  • 6
    Synthesis and applications. Threat criteria provide warnings of population collapse that are consistent with those provided in fisheries stock assessments. Our results suggest that scientists with different backgrounds and objectives should usually be able to agree on the stocks for which the most urgent management action is needed. Moreover, IUCN A1 decline rate metrics may provide useful indicators of population status when the information needed for full fisheries stock assessment is not available.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The understanding of extinction risk in marine organisms is poor, partly as a result of the controversy over whether the methods, originally developed for assessing threat in mammals and birds, can be applied to exploited marine species (Musick 1999; Hutchings 2001a). A number of commercially exploited fishes, including Atlantic cod Gadus morhua L., southern bluefin tuna Thunnus maccoyii (Castelnau 1872) and Atlantic halibut Hippoglossus hippoglossus (L.), were listed as threatened in 1996 under the World Conservation Union (IUCN) Red List criteria because their observed decline rates exceeded the qualifying rate over a time period equal to three generations (Mace & Hudson 1999). Concern was expressed that such listings did not reflect true risk of extinction, especially if the decline was a consequence of management action focusing on maximizing the sustainable yield of a fishery (Mace & Hudson 1999; Powles et al. 2000). Moreover, it was argued that while the IUCN criteria were effective at flagging rapid short-term population changes, they may overestimate extinction risk for many exploited marine species (Matsuda, Yahara & Uozumi 1997; Musick 1999). This weakens the credibility of scientific advice and may result in unnecessarily high short-term political, social and economic costs (Punt 2000). In practice, and notwithstanding high-level political commitment to conserving biodiversity and adopting a precautionary approach, (WSSD 2002), the potentially high costs often weigh against action to conserve biodiversity, particularly when there is any element of uncertainty in the advice (Hilborn 2004; Jennings 2004).

Three studies have explored the validity and utility of threat criteria in marine fishes. Analyses of declines suggest threat criteria thresholds adequately reflect the observed recovery potentials of fish stocks and are consistent with the precautionary approach to fisheries management (Hutchings 2000, 2001a,b). In contrast, two studies suggest the IUCN Red List decline criteria may raise false alarms when applied to commercially exploited fishes. The southern bluefin tuna was listed as critically endangered by the IUCN in 1996, based on an inferred population decline ≥ 80% within three generations (Matsuda et al. 1998). The probability of reaching 500 adults within 100 years was estimated using population viability methods, and found to be 50–96% (Matsuda et al. 1998). It was concluded that southern bluefin tuna was unlikely to become extinct within three to five generations and the 80% decline threshold was too conservative for abundant marine fishes. Punt (2000) compared IUCN Red List decline criteria and fisheries’ reference points using age-structured population projections of six Australian fishes, and found a substantial probability (9–40%) of incorrectly classifying sustainably exploited species as threatened.

We compared threat status and exploitation status for 76 stocks of commercially exploited fishes and invertebrates using three sets of criteria: World Conservation Union (IUCN) population decline criteria A1 and population viability criteria E, and American Fisheries Society, Bethesda, Maryland (AFS) threat criteria. Threat outcomes were compared with an independent assessment of stock exploitation status provided by the International Council for the Exploration of the Sea (ICES). In northern Europe the Advisory Committee on Fisheries Management (ACFM) of ICES categorizes the exploitation status of each stock as ‘inside’ or ‘outside’ safe biological limits by comparing estimated spawning stock biomass and fishing mortality with reference points (ICES 2001; Piet & Rice 2004). We compared threat and exploitation status using a hits, misses and false alarms framework (Rice 2003; Dulvy et al. 2004; Piet & Rice 2004). There are two types of hit: a true positive and a true negative, resulting in a 2 × 2 table of outcomes (Table 1). With this framework: (i) a true positive hit is where a stock is exploited unsustainably (outside safe biological limits) and the threat criteria are met; (ii) a miss is where the species is exploited unsustainably (outside safe biological limits) but does not meet the threat criteria; (iii) a true negative hit is where a stock is exploited sustainably (within safe biological limits) and the threat criteria are not met; and (iv) a false alarm is where the stock is exploited sustainably (within safe biological limits) but the threat criteria are (erroneously) met.

Table 1.  Framework for assessing the performance of threat criteria in relation to the exploitation status of stocks reported by the Advisory Committee on Fisheries Management of the International Council for the Exploration of the Sea (adapted from Rice 2003; Dulvy et al. 2004)
 Stock meets threat criteriaStock does not meet threat criteria
Exploited within safe biological limitsFalse alarmHit (true negative)
Exploited outside of safe biological limitsHit (true positive)Miss

A good set of criteria should avoid false alarms and minimize misses according to signalling theory (Rice 2003). This analysis has wide-ranging implications because it tests for consistency between the advice on population status provided by fisheries and conservation scientists.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

data sources

Time series of numbers at age and spawner–recruit data for 76 commercially exploited North-East Atlantic stocks, comprising 21 species (62 stocks) of fish and three species (14 stocks) of invertebrates, were extracted from ICES stock assessments (http://www.ices.dk). Both sexes were combined in the assessments, except for the Norway lobster Nephrops norvegicus L., where only females were used. Life-history data and additional population information were extracted from published compilations and stock assessments (see Appendix 1). Adult numbers were calculated as the sum of numbers in age classes older than the age at 50% maturity. Body growth rate, k, and asymptotic size, L, were estimated by fitting a von Bertalanffy function to mean lengths at age. Fecundity was calculated as the length at 50% maturity from length–fecundity relationships. Age at 50% maturity was calculated by logistic regression of the proportion of individuals mature at each age. Maximum age was taken as the maximum age used in stock assessments. Generation time was calculated as the midpoint between age at maturity and maximum age. The maximum population growth rate, rmax, was calculated using age at maturity and spawner production (inline image, Myers, Mertz & Fowlow 1997, equation 8). Spawner production (inline image) was estimated by maximum likelihood fit of a Ricker model to spawner–recruit time series (Hilborn & Walters 1992; Myers, Mertz & Fowlow 1997).

categorizing threat

Threat was categorized using three sets of criteria.

IUCN A1

The observed population decline rate was calculated over the greater of 10 years or three generations using time series of adult numbers. Declines of ≥ 90%, ≥ 70% or ≥ 50% meet the IUCN A1 criteria for Critically Endangered, Endangered and Vulnerable. These criteria are appropriate for populations ‘where the causes of the reduction are clearly reversible and understood and ceased’ (IUCN 2004). The A1 criteria have higher thresholds for decline than the A2–4 criteria. One might question how many stocks actually meet these assumptions, but these are the criteria under which most exploited marine fishes are currently assessed (e.g. Cavanagh et al. 2003).

American Fisheries Society

The AFS criteria were used to classify populations or stocks as vulnerable or otherwise using qualifying decline thresholds and a productivity index (Musick 1999). Productivity was categorized as high, medium, low or very low based on any of five life-history parameters (Musick 1999). The maximum population growth rate, rmax, was used to categorize productivity, if available. Otherwise the lowest productivity category calculated using the other life-history traits was used (Musick 1999). The observed population decline rate was calculated over the greater of 10 years or three generations for the corresponding productivity categories. The decline thresholds resulting in a vulnerable listing were 99%, 95%, 85% and 70% for high, medium, low and very low productivities, respectively.

IUCN E

The IUCN E ‘quantitative analysis’ criteria depend on the probability of extinction in the wild over a defined time frame. A population is categorized as Critically Endangered if it exhibits a ≥ 50% probability of extinction within 10 years or three generations, Endangered if it exhibits a ≥ 20% probability of extinction within 20 years or five generations, and Vulnerable if it exhibits a ≥ 10% probability of extinction within 100 years.

We used two population projection approaches to calculate extinction probabilities over the qualifying time frames: non-parametric simulation and diffusion approximation (Dulvy et al. 2004). The probability of extinction was calculated using a non-parametric simulation based on the observed distribution of the rate of change (Matsuda et al. 1998). The rate of change rt was calculated from:

  • image(eqn 1 )

where Nt is the population size in year t. Under the assumption that rt is an independent variable that reflects the probability distribution in each year, then a value for rt can be used to project the current population forwards in time using Nt+1 = Nt exp(rt). This avoids assuming any underlying probability distribution of rt (Matsuda et al. 1998). We calculated the empirical rt values from each interval in time series of adult abundance and projected forward in time from the observed initial population size by bootstrapping rt values with replacement at each time step. A slope was calculated by fitting a least-squares regression model between the simulated population sizes and the observed ones. This was repeated 5000 times and the average slope was calculated. If the average slope was between 0·75 and 1·25, the non-parametric model was judged to provide a reasonable fit to a time series. Stocks not meeting these criteria were not considered further. For the remaining stocks with reasonable model fits, extinction probabilities were calculated from 1000 replicate population projections beginning at the last observed population size. The probability of extinction was calculated as the number of projections attaining the extinction threshold as a proportion of the total number of projections.

Diffusion approximation methods assume that stochastic age-structured populations without density dependence behave as a stochastic discrete time model, where the finite rate of population increase, λ (Dennis, Munholland & Scott 1991, equation 64), is:

  • image(eqn 2 )

The parameters µ and σ2 were estimated using two methods. The Dennis, Munholland & Scott (1991) method estimates µ and σ2 from:

  • image(eqn 3 )
  • image(eqn 4 )

The Holmes (2001) method is relatively insensitive to sampling error, and µ and σ2 are calculated from regression slopes of the mean or variance in running sums of adult abundance over time:

  • image(eqn 5 )
  • image(eqn 6 )

Running sums (Rt) of adult abundance (A) were calculated as:

  • image(eqn 7 )

where L is the number of census counts added to give a running sum at time t, and wi is the weight given to each count. In practice it is difficult to obtain weightings, and so we used a weighting of wi = 1 (Holmes 2001). We used a value of L = 4, which is the maximum possible for time series of the length used here (Holmes 2001). Population size at t + 1 was projected from the initial observed population size from:

  • image(eqn 8 )

using parameters estimated by both the Dennis (Dennis, Munholland & Scott 1991) and Holmes (Holmes 2001) methods. The probability of extinction was calculated only if the Dennis or Holmes methods provided parameter estimates resulting in a reasonable range of projections. Projections were judged to provide a reasonable fit if the average slope of the relationship between the observed and projected abundance was 0·75–1·25.

Extinction risk was calculated as the probability of declining from the last observed population size nq to a lower threshold population size nt from:

  • image(eqn 9 )

using equation 85 in Dennis, Munholland & Scott (1991).

Extinction probabilities were calculated for five future time periods (10 years, 20 years, 3 generations, five generations and 100 years) and six quasi-extinction thresholds (1, 50, 500, 5000, 50 000 and 500 000 individuals), recognizing the possibility that the effective population size is 102−105 times less than the census population size (Hutchinson et al. 2003; Rowe & Hutchings 2003; Dulvy et al. 2004).

categorizing exploitation status

The exploitation status of 68 stocks was extracted from ICES ACFM reports (ICES 2001). Where possible, ACFM defines stocks as exploited ‘inside’ or ‘outside’ safe biological limits. A stock is reported to be ‘outside safe biological limits’ if mean recruitment is lower than if the stock were at its full reproductive capacity (ICES 2001). In order for stocks and fisheries exploiting them to be within safe biological limits, there should be a high probability (i) that the spawning stock biomass is above the threshold where recruitment is impaired (the threshold is defined as Blim) and (ii) that the rate of fishing mortality F is below the threshold that will drive the spawning stock to Blim (Flim). We assumed that the Icelandic spring spawning herring was exploited outside safe biological limits (Jakobsson 1980), resulting in status classifications for a total of 69 stocks.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

iucn a1 criteria

The IUCN A1 criteria could be applied to all 76 stocks, and a total of 12 stocks declined at a sufficient rate to be classified as threatened (Fig. 1; see Appendix 2). The Icelandic spring-spawning herring declined by 99·9% in the last 10 years of the time series and was the only stock to meet the Critically Endangered criteria. Five stocks declined by ≥ 70% over the qualifying time period, meeting the Endangered criteria, including the West Galicia and North Portugal Norway lobster, North Sea cod, East Baltic cod, West Scotland and Ireland herring and Greenland halibut. Six stocks declined by ≥ 70% over the qualifying time period, meeting the Vulnerable criteria, including the Irish Sea cod, Norway coastal cod, West Scotland cod, Baltic sole, West Baltic herring and North Galicia Norway lobster (Fig. 1).

image

Figure 1. Change in adult numbers over time for 12 stocks classified as threatened using the IUCN A1 decline criteria: (a) Icelandic spring spawning herring, Critically Endangered; (b) West Galicia and North Portugal Norway lobster, Endangered; (c) North Sea cod, Endangered; (d) East Baltic cod, Endangered; (e) West Scotland and Ireland herring, Endangered; (f) Greenland halibut, Endangered; (g) Irish Sea cod, Vulnerable; (h) Norwegian coastal cod, Vulnerable; (i) West Scotland cod, Vulnerable; (j) Baltic sole, Vulnerable; (k) West Baltic herring, Vulnerable; (l) North Galicia Norway lobster, Vulnerable. Note the y-axis scales vary. A smoothed loess line (span = 0·25) is presented to indicate the overall trend.

Download figure to PowerPoint

afs criteria

The AFS criteria could be applied to all 76 stocks but only one stock, the Icelandic spring-spawning herring, exhibited the combination of decline rate in adult numbers and productivity to qualify for the Vulnerable category (see Appendix 2). This stock met the vulnerable threshold for the following production metrics: k, fecundity, age at maturity and maximum age. Our results suggested that rmax did not necessarily produce the lowest productivity estimate. Production indices based on k, age at maturity and maximum age were low relative to rmax, and production assignments based on fecundity tended to be higher (see Appendix 2).

iucn e criteria

The non-parametric simulation provided reasonable model fits for 12 stocks. None had a high probability of reaching a population threshold nt ≤ 1 individual within any of the five time scales; hence none met the IUCN E threat criteria (Table 2). Four stocks had a high probability of reaching a population threshold of nt ≤ 500 000 individuals within five generations or 100 years: Kattegat and Norwegian coastal cod, Iberian hake and the North Galician Norway lobster (Fig. 2 and Table 2). The North Galician Norway lobster stock had the highest extinction probability, P = 0·88, of declining below a population threshold of nt ≤ 500 individuals within 100 years.

Table 2.  Extinction probabilities of 12 stocks calculated using the non-parametric simulation method over five time periods and the highest quasi-extinction threshold population size nt = 500 000. Y, year; G, generation
SpeciesStockTime span
10 Y3 G20 Y5 G100 Y
AnchovyBay of Biscay00000·002
AnglerfishCeltic Sea, etc.00000
Atlantic codKattegat000·0020·010·939
Atlantic codNorwegian coastal0000·051
Blue whitingNorthern ICES00000
HakeIberia00000·604
HerringEast Baltic00000
HerringIceland00000
SoleEast English Channel00000
Norway lobsterBotney Gut (North Sea)00000
Norway lobsterFarn deeps (North Sea)00000
Norway lobsterNorth Galicia00·0140·0040·9841
image

Figure 2. Projected change in average adult numbers of four populations over 100 years using the non-parametric simulation approach: (a) North Galicia Norway lobster; (b) Kattegat cod; (c) Norway coastal cod; (d) Iberian hake. The horizontal dotted line represents a quasi-extinction threshold of 500 000 individuals.

Download figure to PowerPoint

The diffusion approximation approach provided reasonable model fits for 12 stocks using the Dennis (Dennis, Munholland & Scott 1991) method and 14 stocks using the Holmes (Holmes 2001) method. Both methods gave similar results when applied to the same stocks (Table 3). Five stocks met the critically endangered criteria, with an extinction probability, P > 0·5, of declining below a population threshold nt < 1 within both 10 years and three generations using the Dennis method: Kattegat cod, Norway coastal cod, Iberian hake, East Baltic herring and North Galician Norway lobster (Table 3). Seven stocks met the critically endangered criteria, with an extinction probability, P > 0·5, of declining below a population threshold nt < 1 within both 10 years and three generations using the Holmes method: Kattegat cod, Norway coastal cod, Northern and Iberian hake and three Norway lobster stocks (North Galician, South-West and South Portugal and Bay of Biscay; Table 3).

Table 3.  Extinction probabilities for 18 stocks calculated using two diffusion approximation methods over five time periods and an extinction threshold of nt = 1. Y, year; G, generation
SpeciesStockDennis, Munholland & Scott (1991) methodHolmes (2001) method
10 Y3 G20 Y5 G100 Y10 Y3 G20 Y5 G100 Y
AnglerfishCeltic Sea, etc.0000000000
Atlantic codKattegat1111111111
Atlantic codNorwegian coastal1111111111
Blue whitingNorthern ICES0000000000
HakeIberia1111111111
HakeNorth Sea, etc.11111
HerringIceland0000000000
HerringEast Baltic11111
HerringGulf of Riga00000
MegrimCeltic Sea, etc.00000
SoleEast English Channel00000
European oysterSolent0·090·080·080·070·04
Norway lobsterBotney Gut00000
Norway lobsterFarn deeps0000000000
Norway lobsterFirth of Forth00000
Norway lobsterNorth Galicia1111111111
Norway lobsterBay of Biscay11111
Norway lobsterSouth-West and South Portugal11111

comparison of threat and exploitation status

A total of 46 (67%) stocks were outside safe biological limits (exploited unsustainably) and 23 (33%) were inside safe biological limits (exploited sustainably). The exploitation status was compared with IUCN A1, AFS and IUCN E threat status. The non-parametric simulation approach was not considered here as none of the stocks assessed with this method met the IUCN E criteria.

None of the threat metrics produced false alarms, i.e. none of the stocks was found to be threatened but was exploited within safe biological limits (Table 4). The quantitative IUCN E metrics both produced higher hit rates and lower miss rates than the decline rate metrics (IUCN A1 and AFS) but could be applied to fewer stocks. The AFS criteria were the most conservative, with the lowest true positive hit rate and the highest true negative hit rate (Table 4).

Table 4.  The proportion (%) of stocks meeting each of four possible outcomes (true positive hit, true negative hit, miss and false alarm) and the total number of stocks for which both stock status and threat status were available. IUCN ED was calculated using the Dennis, Munholland & Scott (1991) method and IUCN EH was calculated using the Holmes (2001) method
Threat criteriaFalse alarmHit (positive)Hit (negative)MissNumber of stocks compared
IUCN A1016364864
AFS0 2366258
IUCN ED036451811
IUCN EH050252512

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Threat criteria and fisheries stock assessments provide comparable information on the status of populations exploited by North-East Atlantic European fisheries. We found no evidence that the application of threat criteria would raise false alarms: none of the threatened stocks was classed as exploited within safe biological limits by ICES. These results lead to two conclusions. First, in every case stocks classified as threatened are unsustainably exploited ‘outside safe biological limits’. Secondly, stocks identified as ‘outside safe biological limits’ may, in some cases, also be considered threatened. Thus management advice to reduce fishing mortality on stocks outside safe biological limits is consistent with the requirement to reduce the risk of extinction, and reductions in fishing mortality will meet the concerns of both fisheries and conservation interests.

Our findings differ from previous analyses suggesting decline criteria may be prone to false alarms when applied to exploited marine species (Matsuda et al. 1998; Punt 2000). Both previous studies evaluated the older IUCN A decline criteria, which had lower decline thresholds, where declines of 20%, 50% and 80% resulted in Vulnerable, Endangered and Critically Endangered categorizations. Also Punt (2000) defined ‘extinction’ as actual extinction, and Matsuda et al. (1998) defined a low quasi-extinction threshold (500 individuals), whereas we used higher quasi-extinction thresholds to account for very low effective population sizes in broadcast spawners (Hauser et al. 2002; Hutchinson et al. 2003; Hoarau et al. 2005). We suspect this change in decline thresholds may underlie the difference in the findings between this and the previous studies; however, we cannot rule out the possibility that the differences may be the result of the different species and modelling approaches used.

Both the decline rate criteria and quantitative criteria indicate that, in addition to the Icelandic spring-spawning herring, 17 overexploited fish and invertebrate stocks may be at increased risk of extinction. These stocks include six cod stocks (North Sea, Irish Sea, East Baltic, West Scotland, Kattegat and Norwegian coast), Iberian and Northern hake, Greenland halibut, Baltic sole, West and East Baltic herring, West Scotland and Ireland herring and four Norway lobster stocks (Bay of Biscay, North Galicia, West Galicia and North Portugal and South-West Portugal). Some of the stocks flagged here have undergone a rapid increase followed by a rapid decrease in abundance, for example Baltic sole and West Scotland and Ireland herring, and threat assessment may be less reliable for strongly fluctuating stocks. The quantitative methods applied here are purely statistical with no mechanistic basis, and therefore the estimate of threat assumes that past management regimes, environmental conditions, habitat availability and quality will also apply in the future.

There are a number of caveats that should be borne in mind when interpreting our results. Marine threat analyses will always involve a trade-off between the degree of threat faced by populations and the availability of data. There are usually few data available for assessing the most threatened marine populations; even when data are available power is too low to detect all except the steepest declines (Dulvy, Sadovy & Reynolds 2003; Maxwell & Jennings 2005). Consequently we were unable to include the most threatened European species in this study, such as populations of common sturgeon Acipenser sturio L., common skate Dipturus batis L., white skate Rostroraja alba (Lacepède 1803), long-nose skate Dipturus oxyrhinchus (L.), angel shark Squatina squatina (L.), tope Galeorhinus galeus (L.), largetooth sawfish Pristis perotteti Müller & Henle 1841, common sawfish Pristis pristis (L.) and Atlantic halibut Hippoglossus hippoglossus (L.) (Dulvy, Sadovy & Reynolds 2003; Baillie, Hilton-Taylor & Stuart 2004). We report results based on a well-understood and consistent marine data set with sufficient variance in threat and exploitation status: 67% of stocks were exploited unsustainably and 24% were threatened. The exclusion of the most threatened species does not undermine this study, indeed most of these threatened species are likely to score as true positive hits, being both unsustainably exploited and threatened.

Maximum age is likely to be underestimated as stock assessments often pool all individuals greater than a certain age, because of the uncertainty of ageing. This negative bias may result in less precautionary (lower) decline rates being measured over a reduced three-generation time window and more precautionary (lower) AFS productivity estimates. The quantitative methods also assume that the variance is solely the result of the population dynamics and demography (Dennis, Munholland & Scott 1991; Holmes 2001). These numbers at age data were derived from Virtual Population Analysis and may be subject to additional variation as a result of observation, process and model error, in addition to the intrinsic variance of the stocks (ICES 2001; Dulvy et al. 2004). Observation errors are known to influence the threat assessment of stocks (Punt 2000), but these are unimportant for our conclusions as the Holmes (2001) method corrects for observation error. However, the model fitting in the stock assessment process should already have eliminated much of the actual observation error in the numbers at age. The quantitative methods used here do not incorporate density dependence. Relatively few stocks can be described by the density-independent diffusion approximation methods, suggesting the dynamics of the other stocks may be better described with density-dependent population viability methods. It is not clear whether this omission leads to overestimation or underestimation of extinction risk. If stocks become more productive in small populations, perhaps because of less competition between individuals for food and space, then the extrapolations into the future will be too pessimistic. Conversely, there is also a very real likelihood of Allee effects (depensation) occurring in marine populations that would lead to greater risk of extinction than indicated by our simulations (Rowe & Hutchings 2003; Dulvy, Freckleton & Polunin 2004; Gascoigne & Lipcius 2004a,b; Rose 2004). Further analyses might explore the effects of model and process errors and incorporate age structure and a range of density-dependent possibilities (Staples, Taper & Dennis 2004).

Our analysis assumes that the ICES exploitation status determinations are ‘correct’. While this may be true for some stocks it will not be true for others, as it is difficult to detect long-term declines in abundance and stock status based on relatively short times series. Threat metrics and ICES exploitation status are also based on the same data and are not strictly independent appraisals of status. We have only compared threat to assessments of exploitation status relative to safe biological limits as defined in a European fisheries management system (ICES 2001a), and there is further scope to compare threat to reference points used in other fisheries management systems.

This analysis supports suggestions that AFS criteria decline rate thresholds are overly conservative and may overlook threatened species (Hutchings 2001a). The AFS criteria assume high resilience in teleost fishes from high interannual variability and reproductive output, but these assumptions have no theoretical or empirical basis (Hutchings 2001a; Sadovy 2001; Denney, Jennings & Reynolds 2002; Hutchings & Reynolds 2004; Myers & Worm 2005). The new CITES criteria use the same decline thresholds as the AFS criteria but are combined with a different decline measure, resulting in even more conservative threat criteria. Marine species have to decline to 5–30% of virgin population sizes before they are considered threatened under the new CITES criteria (FAO 2002; Mace et al. 2002). Our findings suggest that a declines of > 50% over the greater of 10 years or three generations are far in excess of the point of maximum yield and the safe biological limits of many stocks (ICES 2001; Beddington & Kirkwood 2005). In contrast, to meet the new CITES criteria, species must decline even further, by 70–95% from virgin population size, not just over the greater of 10 years or three generations, to qualify as threatened. Consequently the new CITES criteria, like AFS criteria, risk postponing population and species threat listings until their probability of extinction is unduly high and their probability of recovery unduly low.

Concerns about the conservation status of fish stocks have led to the development of regional criteria for assessing extinction risk. For example, the Oslo–Paris Commission for the Protection of the Marine Environment of the North-East Atlantic (OSPAR) has recently adopted the use of a newly developed and relatively unknown set of criteria known as the Texel-Faial criteria (ICES 2001). While the Texel-Faial criteria include a decline criterion, the thresholds are undefined and hence could not be applied to the stocks considered here. Such criteria often have a weaker theoretical grounding than the existing IUCN criteria and, given the apparently reliable performance, international recognition and peer-review of the IUCN criteria, and the consistency with which IUCN criteria reflect ICES fisheries’ management stock assessments, there is little reason to invent new threat criteria without a strong scientific case.

Complex stock assessment models and the IUCN decline criteria both indicate when stocks are exploited unsustainably and may be at risk of extinction. Stock assessments are a financial and logistical impossibility for many target and non-target species that have been depleted by fisheries exploitation, and our results suggest that IUCN criteria provide simple guidance on exploitation status. As this guidance is likely to be consistent with the guidance provided by stock assessments, at least in terms of recommending a direction of change in fishing mortality, the IUCN approach and stock assessment can be regarded as complementary and consistent with an ecosystem approach to fisheries management (Sinclair & Valdimarsson 2003). In management terms, the most important issue is not expected to be the difference between the approaches, but how to reduce fishing mortality effectively and to move exploitation towards sustainability.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank J. Addison, M. Bell, J. Ellis, J. Jakobsson, H. Matsuda, A. Perry, P. Walker and K. Warr for providing data and guidance, and Jake Rice, André Punt and an anonymous referee for their helpful suggestions, and the UK Department of Environment, Food and Rural Affairs (MF0729) and Natural Environment Research Council grant No. # NER/A/S/2001/01075 for funding.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Baillie, J.E.M., Hilton-Taylor, C. & Stuart, S. (2004) 2004 IUCN Red List of Threatened Species: A Global Species Assessment. IUCN, Gland, Switzerland and Cambridge, UK.
  • Beddington, J.R. & Kirkwood, G.P. (2005) The estimation of potential yield and stock status using life-history parameters. Philosophical Transactions of the Royal Society of London, Biological Sciences, 360, 163170.
  • Cavanagh, R.D., Kyne, P.M., Fowler, S.L., Musick, J.A. & Bennett, M.B. (2003) The Conservation Status of Australasian Chondrichthyans. Report of the IUCN Shark Specialist Group Australia and Oceania Regional Red List Workshop. The University of Queensland, School of Biomedical Sciences, Brisbane, Australia.
  • Denney, N.H., Jennings, S. & Reynolds, J.D. (2002) Life history correlates of maximum population growth rates in marine fishes. Proceedings of the Royal Society of London, B, 269, 22292237.
  • Dennis, B., Munholland, P.L. & Scott, J.M. (1991) Estimation of growth and extinction parameters for endangered species. Ecological Monographs, 61, 115143.
  • Dulvy, N.K., Ellis, J.R., Goodwin, N.B., Grant, A., Reynolds, J.D. & Jennings, S. (2004) Methods of assessing extinction risk in marine fishes. Fish and Fisheries, 5, 255275.
  • Dulvy, N.K., Freckleton, R.P. & Polunin, N.V.C. (2004) Coral reef cascades and the indirect effects of predator removal by exploitation. Ecology Letters, 7, 410416.
  • Dulvy, N.K., Sadovy, Y. & Reynolds, J.D. (2003) Extinction vulnerability in marine populations. Fish and Fisheries, 4, 2564.
  • FAO (2002) Report of the Second Technical Consultation on the Suitability of the CITES Criteria for Listing Commercially-Exploited Aquatic Species. Windhoek, Namibia, 22–25 October 2001. Report No. 667. Food and Agriculture Organisation, United Nations, Rome, Italy.
  • Gascoigne, J. & Lipcius, R.N. (2004a) Allee effects driven by predation. Journal of Applied Ecology, 41, 801810.
  • Gascoigne, J. & Lipcius, R.N. (2004b) Allee effects in marine systems. Marine Ecology Progress Series, 269, 4959.
  • Hauser, L., Adcock, G.J., Smith, P.J., Ramirez, J.H.B. & Carvalho, G.R. (2002) Loss of microsatellite diversity and low effective population size in an overexploited population of New Zealand snapper (Pagrus auratus). Proceedings of the National Academy of Sciences, 99, 1174211747.
  • Hilborn, R. (2004) Ecosystem-based fisheries management: the carrot or the stick? Marine Ecology Progress Series (ed. K.I.Stergiou), 274, 275278.
  • Hilborn, R. & Walters, C.J. (1992) Quantitative Fisheries Stock Assessment: Choice, Dynamics and Uncertainty. Chapman & Hall, New York, NY.
  • Hoarau, G., Boon, E., Jongma, D.N., Ferber, S., Palsson, J., Van der Veer, H.W., Rijnsdorp, A.D., Stam, W.T. & Olsen, J.L. (2005) Low effective population size and evidence for inbreeding in an overexploited flatfish, plaice (Pleuronectes platessa L.). Proceedings of the Royal Society, London B, 272, 497503.
  • Holmes, E.E. (2001) Estimating risks in declining populations with poor data. Proceedings of the National Academy of Sciences of the United States of America, 98, 50725077.
  • Hutchings, J.A. (2000) Collapse and recovery of marine fishes. Nature, 406, 882885.
  • Hutchings, J.A. (2001a) Conservation biology of marine fishes: perceptions and caveats regarding assignment of extinction risk. Canadian Journal of Fisheries and Aquatic Sciences, 58, 108121.
  • Hutchings, J.A. (2001b) Influence of population decline, fishing, and spawner variability on the recovery of marine fishes. Journal of Fish Biology, 59A, 306322.
  • Hutchings, J.A. & Reynolds, J.D. (2004) Marine fish population collapses: consequences for recovery and extinction risk. Bioscience, 54, 297309.
  • Hutchinson, W.F. Van Oosterhout, C. Rogers, S.I. & Carvalho, G.R. (2003) Temporal analysis of archived samples indicates marked genetic changes in declining North Sea cod (Gadus morhua). Proceedings of the Royal Society of London, Series B, 270, 21252132.
  • ICES (2001) Report of the ICES Advisory Committee on Fisheries Management, 2001. Report No. 246. International Council for the Exploration of the Sea, Denmark, Copenhagen.
  • IUCN (2004) Guidelines for Using the IUCN Red List Categories and Criteria. World Conservation Union, Gland, Switzerland.
  • Jakobsson, J. (1980) Exploitation of the Icelandic spring and summer-spawning herring in relation to fisheries management. Rapports et Procès-Verbaux des Réunions du Conseil International Pour l’Exploration de la Mer, 177, 2342.
  • Jennings, S. (2004) The ecosystem approach to fishery management: a significant step towards sustainable use of the marine environment. Marine Ecology Progress Series (eds H.I.Browman & K.I.Stergiou), 274, 279282.
  • Mace, G.M. & Hudson, E.J. (1999) Attitudes towards sustainability and extinction. Conservation Biology, 13, 242246.
  • Mace, P.M., Bruckner, A.W., Daves, N.K., Field, J.D., Hunter, J.R., Kohler, N.E., Kope, R.G., Lieberman, S.S., Miller, M.W., Orr, J.W., Otto, R.S., Smith, T.D., Thompson, N.B., Lyke, J. & Blundell, A.G. (2002) NMFS / Interagency Working Group Evaluation of CITES Criteria and Guidelines. NOAA Technical memorandum NMFS-F/SPO-58. US Department of Commerce, Silver Spring, Maryland USA.
  • Matsuda, H., Takenaka, Y., Yahara, T. & Uozumi, Y. (1998) Extinction risk assessment of declining wild populations: the case of the southern bluefin tuna. Researches on Population Ecology, 40, 271278.
  • Matsuda, H., Yahara, T. & Uozumi, Y. (1997) Is tuna critically endangered? Extinction risk of a large and overexploited population. Ecological Research, 12, 345356.
  • Maxwell, D.L. & Jennings, S. (2005) Power of monitoring programmes to detect decline and recovery of rare and vulnerable fish. Journal of Applied Ecology, 42, 2537.
  • Musick, J.A. (1999) Criteria to define extinction risk in marine fishes. Fisheries, 24, 614.
  • Myers, R.A. & Worm, B. (2005) Extinction, survival and recovery of large predatory fishes. Philosophical Transactions of the Royal Society of London, Biological Sciences, 360, 1320.
  • Myers, R.A., Mertz, G. & Fowlow, P.S. (1997) Maximum population growth rates and recovery times for Atlantic cod, Gadus morhua. Fishery Bulletin, 95, 762772.
  • Piet, G.J. & Rice, J.C. (2004) Performance of precautionary reference points in providing management advice on North Sea fish stocks. ICES Journal of Marine Science, 61, 13051312.
  • Powles, H., Bradford, M.J., Bradford, R.G., Doubleday, W.G., Innes, S. & Levings, C.D. (2000) Assessing and protecting endangered marine species. International Council for Exploration of the Seas, Journal of Marine Science, 57, 669676.
  • Punt, A.E. (2000) Extinction of marine renewable resources: a demographic analysis. Population Ecology, 42, 19.
  • Rice, J. (2003) Environmental health indicators. Ocean and Coastal Management, 46, 235259.
  • Rose, G.A. (2004) Reconciling overfishing and climate change with stock dynamics of Atlantic cod (Gadus morhua) over 500 years. Canadian Journal of Fisheries and Aquatic Science, 61, 15531557.
  • Rowe, S. & Hutchings, J.A. (2003) Mating systems and the conservation of commercially exploited marine fish. Trends in Ecology and Evolution, 18, 567572.
  • Sadovy, Y. (2001) The threat of fishing to highly fecund fishes. Journal of Fish Biology, 59A, 90108.
  • Sinclair, M. & Valdimarsson, G. (2003) Responsible Fisheries in the Marine Ecosystem. Food and Agriculture Organisation, Rome, Italy.
  • Staples, D.F., Taper, M.L. & Dennis, B. (2004) Estimating population trend and process variation for PVA in the presence of sampling error. Ecology, 85, 923929.
  • WSSD (2002) Plan of Implementation of the World Summit on Sustainable Development. UN Department of Economic and Social Affairs, Division of Sustainable Development, New York, NY.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Appendix S1. Summary of species, stock area, maximum population growth rate rmax, maximum length L∞ (mm), growth rate, k, fecundity at size of 50% maturity F, age at maturity Tmat, maximum age Tmax and ?na? indicates no data available. Data were extracted from (Beverton 1963; Denney, Jennings & Reynolds 2002; Jakobsson 1980; Jakobsson, Vilhj?lmsson & Schopka 1969; Jennings, Reynolds & Mills 1998; Palmer & Walker 2003; Richardson et al. 1993). Appendix S2. Threat and exploitation status of 76 stocks assessed using IUCN A1 and AFS threat criteria and ICES ACFM exploitation status. Decline rates are based on the greater of 10 year or three-generation time span. Life history traits providing the lowest AFS productivity category are 1 = maximum population growth rate, 2 = body growth rate, 3 = fecundity at length at which 50% of individuals are mature, 4 = age at maturity, and 5 = maximum age. Threat categories of IUCN criteria are Vulnerable (VU), Endangered (EN), Critically Endangered (CR) and not threatened (nt) and AFS criteria are vulnerable (VU), and not threatened (nt). Exploitation status categories are within safe biological limits (w) and outside safe biological limits (o) and uncertain or unknown exploitation status (-).

FilenameFormatSizeDescription
JPE_1063_sm_AppendixS1.doc161KSupporting info item
JPE_1063_sm_AppendixS2.doc170KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.