ICES Journal of Marine Science: Journal du Conseil Advance Access originally published online on January 17, 2007
ICES Journal of Marine Science: Journal du Conseil 2007 64(2):256-270; doi:10.1093/icesjms/fsl032
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Assessing the relative effects of fishing on the New Zealand marine environment through risk analysis
Ministry of Fisheries, PO Box 1020, Wellington 6001, New Zealand
Correspondence to M. L. Campbell: National Centre for Marine and Coastal Conservation, Australian Maritime College, Private Mail Bag 10, Rosebud, 3939, Victoria, Australia; tel: +61 3 5950 2063; fax: +61 3 5981 2158; e-mail: m.campbell{at}ncmcc.edu.au
Campbell, M. L. and Gallagher, C. 2007. Assessing the relative effects of fishing on the New Zealand marine environment through risk analysis – ICES Journal of Marine Science, 64: 256–270.Risk analysis is a tool often used by management to aid decision-making. We present a risk-analysis framework that was developed to facilitate managing New Zealand fisheries. Using catch-effort and observer data, the likelihood that a certain fishery will impact upon five effects of fishing (EoF) issues (non-target species, biodiversity, habitat, trophic interactions, and legislated protected species) is determined. The consequences (impact and/or change) of such events are then determined to determine a relative risk ranking across fisheries. Consequence matrices were developed to assess each of the five EoF categories. To illustrate the model, a 13-y data set of New Zealand fisheries catch-effort and observer data was analysed, using orange roughy (Hoplostethus atlanticus) as an example fishery. The New Zealand fisheries management framework follows a traditional model in which socio-political imperatives are determined (through risk assessment) after ecological impacts are assessed. By maintaining separation between ecological and socio-political imperatives, a transparent and objective framework is established.
Keywords: fisheries management, impact, New Zealand, orange roughy, protected species, relative risk analysis
Received 12 March 2006; accepted 11 November 2006; advance access publication 17 January 2007.
 |
Introduction |
|---|
 Top Introduction Risk analysisApplying the model: orange...DiscussionConclusionsAppendixReferences |
|---|
Changing societal values in the late 20th century have resulted in re-evaluation of past activities, and there has been significant concern about the accelerated degradation of our ecosystems and the apparent inability to assess impacts that activities such as commercial fishing are having on the environment. Many international and national obligations to conserve and protect biodiversity exist [e.g. Convention on the Conservation of Antarctic Living Marine Resources, Convention on Biological Diversity (CBD), Ramsar Convention]. The scientific literature contains many examples of the impacts that fishing can have on the environment (e.g. Carpenter et al., 1985; Thrush et al., 1995; Collie et al., 1997, 2000; Pauly et al., 1998; Tuck et al., 1998; Frid et al., 1999a, 1999b; Hall, 1999; Kaiser et al., 2002, 2006; Queiros et al., 2006), but direct assessment of these impacts is contentious because the analytical tools and data required to prove cause and effect are frequently unavailable. Modelling techniques such as Ecopath, Ecosim, and Ecospace have been suggested as potential exploratory tools (Pauly et al., 1998; Walters, 1999), but a number of limitations have been recognized in applying them to address the ecosystem level consequences of fishing (Robinson and Frid, 2003).
It is invariably argued that effective "ecosystem management" requires the evaluation of a full range of ecosystem "goods and services" (e.g. Lubchenco et al., 1991; Lane and Stephenson, 1998a; Mooney, 1998; Murawski, 2000). However, the practical knowledge to undertake comprehensive ecosystem management is limited by a lack of adequate information and data (biological/ecological and social; Livingston et al., 2005). As such, risk analysis and assessment have proven useful management tools in assessing the biological/ecological aspects of ecosystems when using limited available data (i.e. managing under uncertainty).
In simplistic terms, risk analysis is used to determine how often an event may occur and what the consequences would be of such an event. Within Australia and New Zealand, standards exist that provide best practice for risk management (Australian and New Zealand Standard Risk Management AS/NZ4360:2004; Standards Australia, 2000, 2004). The risk management standard can be summarized in four steps: (i) establishing the context; (ii) identifying the risk; (iii) assessing the risks (risk analysis and risk evaluation); (iv) treating the risks. Recently in Australia, these standards have been used to develop multi-stage risk processes that specifically address wild capture fisheries (Fletcher, 2005; Ye et al., 2005; Astles et al., 2006; Dichmont et al., 2006).
We have developed a semi-quantitative risk-analysis model (herein referred to as the model) to aid New Zealand fisheries science management to: (i) prioritize issues associated with the effects of fishing (EoF) on the environment; (ii) prioritize research needs to ensure that management can operate in a scientifically defensible manner consistent with legislation; (iii) aid the provision of scientific advice to fisheries' managers. The model does not evaluate ecosystem risk directly, but it does identify the relative risks posed by various fisheries. End-point selection becomes critical for effective risk management (Hewitt and Hayes, 2002); the model's end-point assesses the effect (i.e. the impact) commercial fishing has on the environment, specifically non-target species, biodiversity, habitat, trophic interactions, and protected species.
The model is based on ecology and does not analyse socio-political risks posed by fishing; socio-political risks are analysed by fisheries management using a different risk assessment/management process. The model is specifically developed for New Zealand fisheries, in which stocks are managed via a quota system that is single-species focused (i.e. through quotas for individual stocks). On the basis of the risk model's defined end-point, the model does not analyse risk to stock (extinction risk or otherwise). Stock risk is not analysed because the risk model examines the effects that the management regime of a quota stock species has on the rest of the environment (EOF on the environment) relative to other managed fisheries.
Conceptually, within the New Zealand Ministry of Fisheries, risk analysis is undertaken by the science team (i.e. scientific experts), who use the information to provide ecological advice to a Ministries of Fisheries management team (i.e. management experts). The fisheries management team undertakes risk management to ameliorate risk, and hence the advice from science is assessed taking into account social, cultural, and economic issues (Figure 1a). By maintaining separation between risk analysis (science team) and risk management (management team), the Ministry ensures an objective, transparent evaluation of the EOF, removing any socio-political influences from the risk analysis. This follows a traditional management framework for providing fisheries advice and management.
|
An alternative fisheries management framework is presented by Lane and Stephenson (1995, 1998a, 1998b), which emphasizes risk management that maintains risk assessment and risk management within one step and does not delimit risk analysis as a separate step. Hence, biology is only one component of an effective fisheries decision-analysis framework, but this component has equal weighting to social, economic, operational, and industrial components (Figure 1b). By incorporating all components together, any rift between biological advice and other aspects of the problem are given equal weighting. Although the concept has merit, separation of biology from other components is often required in many countries because of signatory obligations to national and international legislation that require specific consideration and reporting on the environment (e.g. NZ Fisheries Act, CBD, Ramsar Convention).
The aim here is to present a semi-quantitative risk-analysis model developed to help assess the relative ecological effects commercial fishing has on the New Zealand marine environment to aid management interventions and to highlight the need for additional data. To illustrate how the model works, we have applied it to commercial fisheries data from the New Zealand orange roughy (Hoplostethus atlanticus) fishery. This fishery was selected randomly as an example only.
|
Risk analysis |
|---|
|
TopIntroductionRisk analysisApplying the model: orange...DiscussionConclusionsAppendixReferences |
|---|
The model requires that all New Zealand fisheries catch-effort and observer data are used in the analyses, but because this is an example to illustrate the risk-analysis process, we have modified and reduced the data set by comparing the orange roughy fishery against all other New Zealand fisheries (555 area-based stocks, representing 93 species). Therefore, we apply the false premise that the modified data set accurately represents all fisheries data in New Zealand. Results presented herein are an example only; they do not represent the true outcome of a risk analysis for the New Zealand orange roughy fishery or the current state of New Zealand fisheries effects on the environment. Moreover, observer data are often patchy and limited, so the assessment of seabird catch is temporally limited by availability of information.
Data
The data used for this analysis are drawn from two New Zealand Ministry of Fisheries databases for catch-effort and observer data. The databases provide the best available New Zealand commercial catch data across all fisheries, but there are limitations. For example, catch-and-effort data are either a record (made at sea) of the top five species caught per tow, ignoring all other species caught per tow, or a record of species landings, which ignores species discards and often concentrates on the main species (landing statistics are demonstrably lower than true catches; see discussion in Hammond and Trenkel, 2004). Similarly, observer data are limited by observer coverage of the fishery (Stratoudakis et al., 1999; Walsh et al., 2002). Further, analyses of each fisheries' catch have often provided information about the top 50 species caught (based on tonnage), ignoring all other species (Anderson et al., 2000; Anderson, 2004). Although these limitations exist, the databases still provide the best available data.
Catch data used for analyses of non-target species, habitat-forming species (HFS), biodiversity taxa, coral taxa (as protected species), and marine mammals (as protected species) come from the catch-effort database and represent all data from 1990 to 2003 for all fisheries. Owing to the limited availability of seabird catch data (limited observer data), the seabird risk analyses are based on data from a draft seabird capture document (2001–2002, i.e. one fishing year; Baird, 2003). Therefore, the seabird catch results presented do not necessarily reflect long-term temporal trends or risks posed to seabirds by commercial fishing, but all other data are long -term, reflecting trends based on a 13-y data set.
Risk-analysis model
For a risk analysis to be effective and transparent, it needs to involve communication, consultation, monitoring, review, and update. Stakeholder and scientific consultation are particularly important in determining percentage values within the consequence matrices (step 3). The model follows the AS/NZ4360:2004 four-step risk-analysis process:
Step 1. Establish the context
The model is used to examine the effects of New Zealand fisheries on the environment, and it can be modified to assess the impacts of introduced species on the environment or expanded to assess economic, social, and cultural risk (Campbell, 2005).
Step 2. Identify the risk, hazards, and effects (i.e. impacts)
We use five broad EoF categories to examine the realized impact of a fishery, categories broadly derived from the literature (Wassenberg et al., 2002; Bundy et al., 2005; Fulton et al., 2005; Mueter and Megrey, 2005), and considered to be the major issues associated with fisheries:
- non-target species, i.e. species of commercial value captured, but which are not target species;
- biodiversity, i.e. all species of non-commercial value captured but not protected or habitat-forming;
- habitat, i.e. habitats that influence fisheries or are impacted by fisheries;
- trophic interactions, i.e. indirect impacts of fishing attributable to flow-on effects on the food chain;
- protected species, i.e. species protected under New Zealand legislation, specifically coral species, marine mammals, and seabirds.
Step 3. Assess the risk (i.e. undertake the risk analysis)
This step is broken into four substeps.
- Determine likelihood. Likelihood is typically described as the probability of an event (impact or incursion) occurring, ranging from rare events to likely or frequent events, and is determined using Table 1. Qualitative and/or quantitative data can feed into this analysis.
- Determine consequence. Consequence measures the impact the fishery may have on the EoF categories. Consequence matrices (Tables 2–6) for each EoF category are used to assess the impact. Separate consequence matrices are used for each category because each may react differently to the impact. For example, a 10% impact upon a protected species that is rare or endangered may be enough to send a protected species extinct. In contrast, a 10% alteration in biodiversity may not be discernible from fluctuations in natural variation (Harwood and Stokes, 2003). The consequence matrices provide multiple examples of varying levels of impact, not all of which are required for that level to be considered relevant.
|
|
|
|
|
|
The consequence matrices were derived via a heuristic process, involving scientific experts (government, industry, and independent scientists) and stakeholders (including Maori, government and industry representatives, conservationists, interested public), in a manner similar to that of Fletcher (2005) and Fletcher et al. (2002, 2004), using working groups and questionnaires. The basis of the threshold values was derived from legislative and policy obligations in the first instance and subsequently adjusted through stakeholder consultation. Consequently, the exact values are subject to adjustment within the constraints of legal and policy frameworks. Threshold values presented here are based on consensus opinion and do not represent a fixed value, but rather a perceived consequence. The use of heuristically derived data is an accepted practice in science and management.
Categories increase in a threshold manner, in which only one example of impact is required to be met to achieve that category. Examples of consequences include the proportion of a fishery's catch, proportion of change, and the number of protected species taken and are not necessarily related to catch (Tables 2–6). A precautionary approach is emphasized, in which a lack of information results in a designation of significant consequence.
- Determine risk. A measure of risk is derived by multiplying likelihood by consequence (Table 7). On the basis of this, a risk ranking is derived. The risk ranking directs action regarding the risk measure (Table 8). The risk-ranking process in the model focuses on scientific action, but this can be altered to include management actions (such as fisheries closures).
- Assess and state uncertainties. Regardless of the method used, evaluations will have uncertainty surrounding the outcomes. This can be due either to measurement error or to real variability in the assessment. Uncertainty exists because there is natural and stochastic variation in environments that is difficult to capture, and mankind has an incomplete understanding of the biological, physical, and anthropogenic systems (Botkin, 1990). This is understandable because ecosystems are highly complex and interconnected, varying both spatially and temporally. It is often impossible to predict ecosystem dynamics (Botkin, 1990; Burgman et al., 1993; Harwood and Stokes, 2003). Uncertainty within this model is discussed because biases relate to reporting of catch data. Data deficiencies were addressed from a conservation perspective, using the precautionary principle (Cooney and Dickson, 2005; Peel, 2005). Therefore, a lack of knowledge is treated as a significant consequence, with the underlying precept that it is better to have a higher probability of Type I error (classify low risk as high risk) than to have high Type II error (classify high risk as low risk), and hence contribute to the collapse of a fishery through poor advice. With improved data reporting and knowledge of habitats and biodiversity, the efficacy of the risk-analysis model will be greatly improved (reducing Type I error). Assessing data deficiencies in such a manner ensures that the analysis is risk averse.
|
|
Step 4. Treat and/or mitigate the risk (if warranted)
On the basis of the risk outcomes, a number of actions can be derived. These actions may involve providing advice to management to treat, ameliorate, or mitigate the risk, taking no action, or directing research. The risk model was developed first to aid science directions and second to provide advice to management. As shown in Figure 1a, the information generated in the risk analysis feeds into fisheries management, so the information is assessed taking into account political imperatives, management goals, and societal values.
The model concentrates on steps 2 and 3, because step 1 is already defined and step 4 is typically covered under risk assessment (management process). Because it is a semi-quantitative model (using real data collected in a quantifiable manner to identify categorical assessments), it can incorporate the outcomes of deterministic or stochastic models to provide results for the likelihood and consequence matrices if required. As stated earlier, the end-point of the risk analysis is to assess the relative impact a New Zealand fishery has upon the environment. This end-point reflects a need, under New Zealand fisheries legislation and international obligations, to evaluate management regimes for fisheries in relation to protection of the environment.
Analysis can take place at the level of each of the five EoF categories or at subcomponent levels within each EoF category. For example, each protected species group (as defined under national legislation; e.g. seabirds, marine mammals, and corals) is a subcomponent of the protected species EoF category, and each group undergoes separate risk assessments.
|
Applying the model: orange roughy fishery |
|---|
|
TopIntroductionRisk analysisApplying the model: orange...DiscussionConclusionsAppendixReferences |
|---|
Step 1. Establish the context
The New Zealand fishery for orange roughy is predominantly a deep-water trawl fishery (99.88% of total catch), with just limited effort being expended/recorded in midwater trawl (0.12% of total catch). In New Zealand, the fishery operates in depths of 600–1500 m across eight regions, termed fisheries management areas (FMAs). Total allowable commercial catch (TACC) has declined from 38 000 t in the 1991/1992 fishing year to the current TACC of 21 114 t. TACCs are altered annually to reflect the principles of sustainable stock management, the orange roughy stock having been fished down to close to the level of biomass that will support maximum sustainable yield (BMSY).
Catch-effort data from the 13-y data set demonstrate that, on average across the annual orange roughy fishery, catch was 
23 906 t, representing 5.6% of the total catch of all 93 target species commercial fisheries in New Zealand. About 1158 t of orange roughy is caught annually as bycatch in other commercial fisheries, such as those targeting oreo (e.g. smooth oreo, Pseudocyttus maculatus), cardinalfish (Epigonus sp.), hoki (Macruronus novaezelandiae), rubyfish (Plagiogeneion rubiginosum), alfonsino (Beryx splendens), and gemfish (Rexea solandri).
Th fishery operates in areas typically associated with seamounts. Seamounts are considered to be particularly vulnerable ecosystems that contain many locally endemic species (Koslow et al., 2000). Because of the linkage and perceived risk between the fishery and the vulnerable seamount ecosystems, the orange roughy fishery has drawn particular attention from conservation groups over the past decade (Johnston and Santillo, 2002). To date, orange roughy fisheries around the globe have been linked with the destruction of protected species (Probert et al., 1997) and rapid declines in stock (Clark, 1999a, 1999b).
Step 2. Identify the risks, hazards, and effects
As stated earlier, five fishing-effects categories were identified as issues that needed to be assessed. The categories were drawn from the literature, and they provide a consistent foundation against which the EoF can be assessed, while also meeting national and international obligations.
Step 3. Assess the risk
The risk results for the orange roughy fishery are summarized in Table 9. The likelihood of collecting non-target species during orange roughy trawling was rated as likely, catches of 169 non-target species (including species complexes) being recorded. The relative consequence of this catch is minor in relation to other fisheries, however, given that the catch of non-target species in the fishery constitutes just 5.7% of the non-target species caught over the 93 targeted commercial fisheries examined. Therefore, given the combination of likely and minor, a moderate relative-risk ranking is generated (Table 9). Scientific actions (such as improved research into orange roughy bycatch) could be taken to reduce this risk ranking.
|
The likelihood of this deep-water trawl fishery impacting habitat is also likely. The orange roughy fishery has reported a catch of three habitat-forming species (HFS) (Figure 2), representing 0.02% of the total HFS take reported across all fisheries examined (Figure 3). This information, although important, is not utilized in the consequence matrix: the consequence matrix requires information pertaining to change in habitat type in the fishing areas, or HFS population data. Therefore, although data exist on the catch of HFS, there is a lack of information about habitat type and HFS population data. In general, there are few baseline data pertaining to habitats and HFS in New Zealand fishery areas. To account for this lack of knowledge in a conservative (precautionary) fashion, and based on the habitat consequence matrix, a rating of significant consequence is the default. The outcome of this is a relative risk ranking of extreme (Table 9). The risk ranking for this fishery can be reduced if more information about habitats, their distributions, and/or the habitats or HFS composition can be obtained.
|
|
The likelihood of the orange roughy fishery impacting biodiversity is considered likely, with 11 biodiversity taxa reported as bycatch (Figure 2). The consequence of collecting these taxa is insignificant given that the catch represents 0.2% of the total biodiversity catch reported across all examined fisheries (Figure 3). Therefore, a relative-risk ranking of negligible is applied because the risk is considered to be zero (Table 9). Although this ranking does not require scientific action, because of the bottom-trawl methods utilized in the fishery, it is likely that biodiversity is affected (Probert et al., 1997; Tuck et al., 1998; Clark, 1999b; Clark et al., 2000; Koslow et al., 2000; Jennings et al., 2001; Kaiser et al., 2006; Queiros et al., 2006); so it is suggested that reporting of biodiversity species within the fishery be verified.
Trophic interactions for all fisheries now cannot be assessed owing to an inability to determine the likelihood of such an event (Table 9). If such an event occurred, we could estimate that the result would be significant, but without likelihood, there can be no risk ranking. Further research within New Zealand waters is required to determine the likelihood of trophic interactions.
The likelihood of the orange roughy fishery impacting on protected species is rare for marine mammals and seabirds, but likely for corals. Both marine mammals (n = 2; Figure 2) and seabirds have been reported as catch in the fishery. Although likelihood is rare, the consequence of marine mammal catch is major, a high 11.4% of the total catch of marine mammals across all fisheries (Figure 3). The resultant relative risk ranking is moderate, based on the combination of rare and major (Table 9). To mitigate this risk, scientific action should be implemented to reduce the catch of protected species.
The consequence of seabird catch in the orange roughy fishery is minor, just 0.4% of the catch from all fisheries in the period 2001–2002 (Figure 3). Observer data for the same period show that white-capped albatross (Thalassarche steadi) and southern royal albatross (Diomedea epomophora) were the only species taken as bycatch in this fishery (Baird, 2003). In New Zealand and Australia, both species are classified as vulnerable (IUCN categorization), facing a high risk of extinction in the wild in the medium-term future. Genetically, the shy albatross (T. cauta) has been demonstrated to have evolved from the widely dispersed white-capped albatross (Abbott and Double, 2003a, 2003b), so its distribution is potentially more widespread than previously thought. Observer coverage of the fishery is low and information regarding catch of seabirds may increase with improved observer coverage. The consequence is considered minor, resulting in a low relative-risk ranking, but given that the species may not recover from even minor impact, the consequence outcome is uncertain (Table 9). If the null hypothesis tested is that fishing undertaken by the orange roughy fishery does not impact the seabird population (although observer coverage is limited), then acceptance of the null hypothesis when it is incorrect may occur (Type II error). Consequently, increased science action is required to obtain further observer data.
The likelihood of the orange roughy fishery impacting protected corals is likely. The consequence of coral catch in this fishery is significant, a high 98% of the total coral catch by all fisheries examined (Figures 2 and 3). The resultant relative risk ranking is extreme, owing to the combination of likely and significant (Table 9), so prompt scientific action is required.
Step 4. Treat and/or mitigate the risk
Science can provide a number of actions that could be implemented to ameliorate or mitigate risk in the analyses. Examples of management action that could ameliorate effects include reducing TACC or restricting trawling activities within and across FMAs, increasing observer coverage to increase data accuracy, developing industry codes of practice, and developing risk maps. Scientific actions could include research to reduce bycatch and instigation of improved monitoring and data recording to ensure accurate representation of the data.
Improved fishing codes of practice and the National Plan of Action for Seabirds may result in reduced seabird mortality (S. Waugh, pers. comm.), and improved fishing technologies may result in better marine mammal exclusion devices. The Hoki Fishery Management Company and the Squid Fishery Management Company have implemented codes of practice aimed at reducing the incidental catch of seabirds (S. Clubb, pers. comm.; http://www.mfish.govt.nz/sustainability/impacts.html). On the basis of the limited data available from Baird (2003), incidental catch of seabirds in the hoki fishery ranks fourth in New Zealand behind the ling, squid, and tuna fisheries. Similarly, the squid fishery has developed a code of practice to reduce catch of marine mammals, which are protected under New Zealand legislation. That fishery currently accounts for 10.9% (fifth-ranked fishery in order of reported catch) of all marine mammals caught in New Zealand fisheries (unpublished catch-effort data), with a reported marine mammal catch over the 13-y data-reporting period consisting of fur seals, Hooker's sea-lions, seals, and dusky dolphins.
Improved fishing methods will reduce bycatch and the risk that fisheries pose to the environment. Such improvements take place through application, incentives, or technology. Globally, there has been much expenditure and research on developing non-target species exclusion devices. Obvious examples include the turtle exclusion devices (TEDs) used in Northern Australia (Salini et al., 2000) and the sea-lion exclusion devices (http://www.mfish.govt.nz/sustainability/management-strategy/nzSeaLionOperationalPlan-0605.pdf) used in New Zealand. Bycatch reduction devices, along with appropriate net sizing, are also used to increase the escape rate of non-target species from trawls (Garcia-Caudillo, 2000; Salini et al., 2000; Criales-Hernandez et al., 2006), without significantly reducing target species catch.
The risk model is readily applicable to the development of risk maps through a spatially explicit Geospatial Information System (GIS). As a risk management tool, the use of a spatially explicit GIS to map habitats and species with species and/or fishing method likelihood and consequence will result in the ability to produce risk maps. For this to take place in New Zealand, data on the habitats within each FMA are required. Currently, a broad-scale resolution (0.10 degrees) is suggested, delimiting the following habitats: reef (coral, bryozoan, molluscs, artificial, and so on), seamount, sand, mud, seagrass, kelp, and rubble. Information on habitat resilience would also be of benefit for these analyses. The potential outcomes of the use of risk maps are that habitat types for each FMA would be represented, with changes in habitat risk map communities being both spatially and temporally monitored. To a limited extent, this is already being undertaken by the Ministry of Fisheries through the NABIS (National Aquatic Biodiversity Information System) project (http://www.biodiversity.govt.nz/seas/biodiversity/programmes/nabis/index.html).
|
Discussion |
|---|
|
TopIntroductionRisk analysisApplying the model: orange...DiscussionConclusionsAppendixReferences |
|---|
We used a semi-quantitative risk-analysis model to measure the relative risks of environmental effects between New Zealand fisheries. The risk analysis developed concentrates on non-target species, habitat, biodiversity, trophic interactions, and protected species and, in the example provided, analyses the risk against other fisheries, not the natural background conditions (which can be implemented as discussed subsequently). The use of risk analysis as a fisheries management tool has become more common, with recent work by Fletcher et al. (2002, 2004), Fletcher (2005), and Astles et al. (2006) being implemented in the Australian context. The model created by Fletcher and colleagues builds on stakeholder input and relies on qualitative data and social and economic values.
It is generally accepted that a precautionary approach to sustainable development is needed (de la Mare, 2005), and as such fisheries management in New Zealand attempts to meet its CBD obligations. The relative risk-analysis process outlined here uses a precautionary approach: identifying where there is little or no information as significant consequence outcomes. The model also applies some of the precepts of sustainable development by:
- identifying undesirable outcomes (impacts on the environment);
- providing a further decision tool (risk analysis and risk management) for a legal and institutional framework;
- using an extensive data set, which can account for intergenerational impacts and feed into the needs of the future.
Application of the model highlights uncertainty in the data associated with information gaps. For example, biodiversity and habitat data deficiencies are identified, inferring inadequacies in the catch-effort data that may bias analyses. The model uses a conservative approach when faced with limited data, so if the restricted data set is acknowledged at the outset, a conservative ranking can be accorded, a method favoured by the CBD. However, if the limitations of the data are not reported at the outset, then the risk assessment outcomes may suffer from Type II error (accepting low risk when in fact risk is high). For example, limited observer coverage will potentially lead to Type II error in seabird and marine mammal risk analyses. In general, it needs to be acknowledged that catch-effort and observer data are uni-directionally biased, i.e. the reported species mix and tonnage represent the minimum catch when more species/tonnage may have been caught, but not reported for a variety of reasons (e.g. reporting format, training, observer coverage, catch non-reporting).
From the outcomes of our model, we deduce that improvements in the accuracy and consistency of catch reporting will ensure more realistic evaluation and reduce Type II error. For example, landings data do not record all species collected, and data collected at sea only represent the top five (by weight) species caught. Hence, valuable biodiversity and habitat data are typically not available for the evaluation. This situation needs to be rectified in a pragmatic manner (considering both environmental and economic imperatives) to ensure that impacts on biodiversity and habitat are accurately recorded and effective management actions instigated.
In the example fishery assessed here (for orange roughy), the model determined the impacts that the fishery had on non-target species, habitats, biodiversity, and protected species when compared with other New Zealand fisheries. However, analysis against natural background conditions was not demonstrated because of the data set used (catch-effort). To provide a more robust analysis, inclusion of natural background conditions is facilitated by the use of scientific trawl data, which provides some data on natural background conditions as well as providing information used to assess stocks.
Data limitations, whether related to scientific trawl, catch-effort, or observer data, are affected by illegal, unreported, and unregulated (IUU) fishing. A number of stock models and risk-analysis methods for IUU in southern waters are currently being developed. However, they make the assumption that IUU fishing operates in the same way as regulated fishing, which is not a conservative approach. IUU fishers have already shown a disregard for rules, and so it would be more conservative to consider that IUU fishing does not follow standards/accepted fishing practice (e.g. disregarding net sizes, TED requirements). Therefore, although the model attempts to determine risk, a more accurate risk assessment would also need to include estimated IUU fishing contributions.
The model was unable to assess the impact on trophic interactions because likelihood could not be determined. Currently, a number of projects are attempting to fill this information gap. However, the model does identify mitigation options, such as improvements in fishing methods and data collection, data recording, and data reporting.
For the outputs of the model to be effective, they need to feed into a complete risk assessment process. Risk assessment evaluates the scientific advice provided by the model, incorporating real and perceived values with socio-political imperatives. Within New Zealand, socio-political imperatives are placed in the context of four "core values": environmental (covered by this model), economic (fishery market value, infrastructure), social (e.g. aesthetics, amenity), and cultural (e.g. Maori and iconic values) (Ministry for the Environment, 1995). Through analysis of all four values, a pragmatic management decision can be made while maintaining an objective, transparent, and distinct scientific evaluation.
|
Conclusions |
|---|
|
TopIntroductionRisk analysisApplying the model: orange...DiscussionConclusionsAppendixReferences |
|---|
While target fish stocks are managed, the effects fishing activities have on the marine environment are unknown and/or equivocal. Given this lack of information, management recognizes a need to improve its understanding and to focus limited resources to improve management outcomes. An evaluation of the relative risks posed by managed fisheries to the marine environment can help prioritize issues and research needs in a cost-effective manner.
As with all risk-analysis processes, the outcomes are only as good as the data input. The strength of the model is its ability to assess risk in a semi-quantitative fashion while highlighting gaps in which improved data collection is required and treating data deficiencies in a conservative fashion. The model follows a standard risk-analysis process: determining likelihood and consequence and using these values to derive risk. The process can be applied to produce risk maps for regions or FMAs, which again will aid fisheries management with positive flow-on effects for the environment.
|
Appendix |
|---|
|
TopIntroductionRisk analysisApplying the model: orange...DiscussionConclusionsAppendixReferences |
|---|
Only codes used in Figures 2 and 3 are provided. The codes are from New Zealand's catch-effort database. Additional codes can be obtained from the Research Data Management Group of New Zealand's Ministry of Fisheries (PO Box 862, Wellington, New Zealand).
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
aThis designation does not exist in the catch-effort database, but represents a summary of all tuna fisheries.
|
Acknowledgements |
|---|
This work was undertaken by MLC while working at the Ministry of Fisheries, New Zealand. We thank the following people for providing guidance on fisheries stock assessment and the New Zealand QMA process: Chris O'Brien, Marc Griffiths, Merryn Jones, Pamela Mace. MLC also thanks Chad Hewitt (National Centre for Marine and Coastal Conservation, Australia) and Dan Lane (University of Ottawa, Canada) for useful discussions on risk management frameworks. Finally, we acknowledge the efforts of the two reviewers whose comments strengthened the final manuscript.
|
References |
|---|
|
TopIntroductionRisk analysisApplying the model: orange...DiscussionConclusionsAppendixReferences |
|---|
-
Abbott C. L. and Double M. C. (2003a) Phylogeography of shy and white-capped albatrosses inferred from mitochondrial DNA sequences: implications for population history and taxonomy. Molecular Ecology 12:2747–2758.[CrossRef][Medline]
Abbott C. L. and Double M. C. (2003b) Genetic structure, conservation genetics and evidence of speciation by range expansion in shy and white-capped albatrosses. Molecular Ecology 12:2953–2962.[CrossRef][Medline]
Anderson O. (2004) Fish discards in the trawl fisheries for arrow squid, jack mackerel, and scampi in New Zealand waters. (Ministry of Fisheries, New Zealand)48 New Zealand Fisheries Assessment Report.
Anderson O., Clark M., Gilbert D. (2000) Bycatch and discards in trawl fisheries for jack mackerel and arrow squid, and in the longline fishery for ling, in New Zealand waters. Final Research Report. (Prepared for the Ministry of Fisheries, New Zealand)44 NIWA Technical Report, NTR74.
Astles K. L., Holloway M. G., Steffe A., Green M., Ganassin C., Gibbs P. J. (2006) An ecological method for qualitative risk assessment and its use in the management of fisheries in New South Wales, Australia. Fisheries Research 82:290–303.[CrossRef]
Baird S. J. (2003) Estimation of the incidental capture of seabird species in commercial fisheries in New Zealand waters, 2001–02. (Ministry of Fisheries, New Zealand)61 Draft New Zealand Fisheries Assessment Report September.
Botkin D. B. (1990) Discordant Harmonies: a New Ecology for the Twenty-First Century. (Oxford University Press, New York)241.
Bundy A., Fanning P., Zwanenburg K. 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.
Burgman M. A., Ferson A., Akcakaya H. R. (1993) Risk Assessment in Conservation Biology. (Chapman and Hall, London)314.
Campbell M. L. (2004) Ecological risk assessment of the effects of fishing on the environment: a tool to aid scientific advice to fisheries managers. 142 Unpublished Internal Ministry of Fisheries Document.
Campbell M. L. (2005) A Draft Guide for Risk Analysis Applied to Alien Marine Species (pertaining to the Barcelona Convention). , Tunisia40 All Oceans Ecology Report (AOE2005-4) prepared for Regional Activity Centre for Specially Protected Areas (RAC SPA) of the UNEP Mediterranean Action Plan.
Carpenter S. R., Kitchell J. F., Hodgson J. R. (1985) Cascading trophic interactions and lake productivity. BioScience 35:634–639.[CrossRef][ISI]
Clark M. (1999a) Are deepwater fisheries sustainable? – the example of orange roughy (Hoplostethus atlanticus) in New Zealand. Fisheries Research 51:123–135.[CrossRef]
Clark M. (1999b) Fisheries for orange roughy (Hoplostethus atlanticus) on seamounts in New Zealand. Oceanologica Acta 22:593–602.[Medline]
Clark M. R., Anderson O. F., Francis R. I. C. C., Tracey D. M. (2000) The effects of commercial exploitation on orange roughy (Hoplostethus atlanticus) from the continental slop of the Chatham Rise, New Zealand, from 1979 to 1997. Fisheries Research 45:217–238.[CrossRef]
Collie J. S., Escanero G. A., Valentine P. C. (1997) Effects of bottom fishing on the benthic megafauna of Georges Bank. Marine Ecology Progress Series 155:159–172.[ISI]
Collie J. S., Escanero G. A., Valentine P. C. (2000) Photographic evaluation of the impacts of bottom fishing on benthic epifauna. ICES Journal of Marine Science 57:987–1001.
Cooney R. and Dickson B. (2005) Biodiversity and the Precautionary Principle. Risk and Uncertainty in Conservation and Sustainable Use. (Earthscan, London)314.
Criales-Hernandez M. I., Duarte L. O., Garcia C. B., Manjarrés L. (2006) Ecosystem impacts of the introduction of bycatch reduction devices in a tropical shrimp trawl fishery: insights through simulation. Fisheries Research 77:333–342.[CrossRef]
de la Mare W. K. (2005) Marine ecosystem-based management as a hierarchical control system. Marine Policy 29:57–68.[CrossRef]
Dichmont C. M., Deng A., Punt A. E., Venables W., Haddon M. (2006) Management strategies for short-lived species: the case of Australia's northern prawn fishery. 1. Accounting for multiple species, spatial structure and implementation uncertainty when evaluating risk. Fisheries Research 82:235–245.[CrossRef]
Fletcher W. J. (2005) The application of qualitative risk assessment methodology to prioritize issues for fisheries management. ICES Journal of Marine Science 62:1576–1587.[CrossRef]
Fletcher W. J., Chesson J., Fisher M., Sainsbury K. J., Hundloe T. J. (2004) National ESD Reporting Framework: The ‘How To. , Canberra88 Guide for Aquaculture, version 1.1. FRDC.
Fletcher W. J., Chesson J., Fisher M., Sainsbury K. J., Hundloe T., Smith A. D. M., Whitworth B. (2002) National ESD Reporting Framework for Australian Fisheries: The ‘How To’ Guide for Wild Capture Fisheries. , Canberra120 FRDC Project, 2000/145.
Frid C. L. J., Clark R. A., Hall J. A. (1999a) Far-field control of long-term changes in Northumberland (NW North Sea) coastal zooplankton. ICES Journal of Marine Science 53:972–977.
Frid C. L. J., Hansson S., Ragnarsson S. A., Rijnsdorp A., Steingrimsson S. A. (1999b) Changing levels of predation on benthos as a result of exploitation of fish populations. Ambio 28:578–582.
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-Caudillo J. M., Cisneros-Mata M. A., Balmori-Ramírez A. (2000) Performance of a bycatch reduction device in the shrimp fishery of the Gulf of California, México. Biological Conservation 92:199–205.[CrossRef]
Hall S. J. (1999) The Effects of Fishing on Marine Ecosystems and Communities. (Blackwell Science, Oxford)296 Fish Biology and Aquatic Resources Series 1.
Hammond T. R. and Trenkel V. M. (2004) Censored catch data in fisheries stock assessment. ICES Journal of Marine Science 61:274–286.
Harwood J. and Stokes K. (2003) Coping with uncertainty in ecological advice: lessons from fisheries. Trends in Ecology and Evolution 18:617–622.[CrossRef]
Hewitt C. L. and Hayes K. R. (2002) Risk assessment of marine biological invasions. In Leppakoski E., Gollasch S., Olenin S. (Eds.). Invasive Aquatic Species of Europe(Kluwer Academic, Dordrecht) pp. 456–466 583.
Jennings S., Dinmore T. A., Duplisea D. E., Warr K. J., Lancaster J. E. (2001) Trawling disturbance can modify benthic production processes. Journal of Animal Ecology 70:459–475.[CrossRef][ISI]
Johnston P. and Santillo D. (2002) Conservation of seamount ecosystems: application of an MPA concept. Archive of Fishery and Marine Research 51:305–319.
Kaiser M. J., Clarke K. R., Hinz H., Austen M. C. V., Somerfield P. J., Karakassis I. (2006) Global analysis of response and recovery of benthic biota to fishing. Marine Ecology Progress Series 311:1–14.
Kaiser M. J., Collie J. S., Hall S. J., Jennings S., Poiner I. R. (2002) Impacts of fishing gear on marine benthic habitats. Reykjavik Conference on Responsible Fisheries in the Marine Ecosystem. FAO Fisheries Report 658:197–217.
Koslow J. A., Boehlert B. W., Gordon J. D. M., Haedrich R. L., Lorance P., Parin N. (2000) Continental slope and deep-sea fisheries: implications for a fragile ecosystem. ICES Journal of Marine Science 57:548–557.
Lane D. E. and Stephenson R. L. (1995) Fisheries management science: the framework to link biological, economic and social objectives in fisheries management. Aquatic Living Resources 8:215–221.
Lane D. E. and Stephenson R. L. (1998a) A framework for risk analysis in fisheries decision-making. ICES Journal of Marine Science 55:1–13.
Lane D. E. and Stephenson R. L. (1998b) Fisheries co-management: organization, process, and decision support. Journal of Northwest Atlantic Fishery Science 23:251–265.
Livingston P. A., Aydin K., Boldt J., Ianelli J., Jurado-Molina J. (2005) A framework for ecosystem impacts assessment using an indicator approach. ICES Journal of Marine Science 62:592–597.
Lubchenco J., Olson A. M., Brubaker L. B., Carpenter S. R., Holland M. M., Hubbell S. P., Levin S. A. (1991) The sustainable biosphere initiative: an ecological research agenda. Ecology 72:371–412.[CrossRef]
Ministry for the Environment. (1995) Environment 2010 Strategy – A Statement of the Government's Strategy on the Environment. (Ministry for the Environment, Wellington, New Zealand)63.
Mooney H. A. (1998) Ecosystem management for sustainable marine fisheries. Ecological Applications 8:S1.[CrossRef][ISI]
Mueter F. J. and Megrey B. A. (2005) Distribution of population-based indicators across multiple taxa to assess the status of Gulf of Alaska and Bering Sea groundfish communities. ICES Journal of Marine Science 62:344–352.
Murawski S. A. (2000) Definitions of overfishing from an ecosystem perspective. ICES Journal of Marine Science 57:649–658.