Seafood Watch Criteria for Fisheries Review by Trevor J Ward

Version October 22, 2014

Seafood Watch Criteria for Fisheries

Please include your contact details below All documents submitted during the public consultation process will be posted on our website. Documents will posted exactly we receive them except that this front page will be removed. The organization/author field below will be displayed as the ‘author’ of the online posted document. If you wish for your document to remain anonymous, please indicate in the check box below. If ‘Anonymous’ is selected, the ‘author’ of this document when posted on our website will then simply read ‘Anonymous.’

Organization/Author Trevor J Ward, University of Technology Sydney, Australia Name of point person Email [email protected] Click if you would you like to remain anonymous ☐

1 Version October 22, 2014

Seafood Watch® Criteria for Fisheries

Public comment period – 1

Introduction The Monterey Bay Aquarium is requesting and providing an opportunity to offer feedback on the Seafood Watch Fisheries Assessment Criteria during our current revision process. Before beginning this review, please familiarize yourself with all the documents available on our Standard review website.

Providing feedback, comments and suggestion This document contains the current version of the Seafood Watch criteria. “Guidance for public comment” sections have been inserted and highlighted, and various general and specific questions have been asked throughout. In certain sections, proposed changes have been noted using redlining for proposed new text and strikethrough for proposed deletion. Narrative descriptions and rationale for proposed changes are also described in the respective comment sections. Seafood Watch welcomes feedback and particularly suggestions for improvement on any aspect of the criteria. Please provide feedback, supported by references wherever possible in any sections of the criteria of relevance to your expertise.

Seafood Watch Ratings The Seafood Watch Criteria for Fisheries are used to produce assessments for wild-capture fisheries resulting in a Seafood Watch rating of Best Choice (green), Good Alternative (yellow), or Avoid (red). The assessment criteria are used to determine a final numerical score as well as numerical subscores and color ratings for each criterion. These scores are translated to a final Seafood Watch color rating according to the methodology described in the table below. The table also describes how Seafood Watch defines each of these categories. The narrative descriptions of each Seafood Watch color rating category, and the guiding principles listed below, compose the framework the criteria are based on, and should be considered when providing feedback on any aspect of the criteria.

Best Choice Final Score >3.21, and no Red Wild-caught and farm-raised seafood on the “Best Criteria, and no Critical2 scores Choice” list are ecologically sustainable, well managed

1 Each criterion is scored from 1 to 5 based on subfactor scores, as described in the document below. Criteria scoring <= 2.2 are considered “red” criteria. The final numerical score for the fishery is calculated as the geometric mean of the four Scores (Criterion 1, Criterion 2, Criterion 3, Criterion 4). 2 Very severe conservation concerns receive “critical” scores, which result in an Avoid recommendation.

2 Version October 22, 2014

and caught or farmed in ways that cause little or no harm to habitats or other wildlife. These operations align with all of our guiding principles. Good Final score >2.2, and neither Harvest Wild-caught and farm-raised seafood on the “Good Alternative Strategy (Factor 3.1) nor Bycatch Alternative” list cannot be considered fully sustainable at Management Strategy (Factor 3.2) this time. They align with most of our guiding principles, are Very High Concern3, and no but there is either one conservation concern needing more than one Red Criterion, and no substantial improvement, or there is significant Critical scores, and does not meet uncertainty associated with the impacts of this fishery or the criteria for Best Choice (above) aquaculture operations. Avoid Final Score <=2.2, or either Harvest Wild-caught and farm-raised seafood on the “Avoid” list Strategy (Factor 3.1) or Bycatch are caught or farmed in ways that have a high risk of Management Strategy (Factor 3.2) is causing significant harm to the environment. They do not Very High Concern5, or two or more align with our guiding principles, and are considered Red Criteria, or one or more Critical unsustainable due to either a critical conservation scores. concern, or multiple areas where improvement is needed.

The SFW Assessment Principles

1. Follow the principles of ecosystem-based fisheries management The fishery is managed to ensure the integrity of the entire ecosystem, rather than solely focusing o maintenance of single species stock productivity. To the extent allowed by the current state of the science, ecological interactions affected by the fishery are understood and protected, and an appropriate level of precaution is applied in management decisions to reflect uncertainty in the knowledge base, to maintain the structure and function of the ecosystem that may be influenced by fishing. 2. Ensure all affected stocks4 are healthy and abundant Abundance, size, sex, age and genetic structure of the main species affected by the fishery (not limited to target species) is maintained at levels that do not impair recruitment or long-term productivity of the stocks or fulfillment of their role in the ecosystem and food web. Abundance of the main species affected by the fishery should be at, above, or fluctuating around levels that allow for the long-term production of maximum sustainable yield if that level is consistent with maintaining healthy structure and function of the ecosystems where such species may live at the different stages of their life cycle. 3. Fish all affected stocks at sustainable levels Fishing mortality for the main species affected by the fishery should be appropriate to at least maintain MSY given current abundance and inherent resilience to fishing while accounting for scientific uncertainty, management uncertainty, and non-fishery impacts such as habitat degradation by applying levels of precaution that are appropriate to protect ecosystem structure and function. The cumulative fishing mortality experienced by affected species must be at or below the level that produces maximum sustainable yield for single-species fisheries on typical species that are at target levels.

3 Because effective management is an essential component of sustainable fisheries, Seafood Watch issues an Avoid recommendation for any fishery scored as a Very High Concern for either factor under Management (Criterion 3). 4 “Affected” stocks include all stocks affected by the fishery, no matter whether target or bycatch, or whether they are ultimately retained or discarded.

3 Version October 22, 2014

Fishing mortality may need to be lower than the level that produces maximum sustainable yield in cases such as multispecies fisheries, highly vulnerable species, or fisheries with high uncertainty. For species that are depleted below target levels, fishing mortality must be at or below level that allows the species to recover to its target abundance within 5 years. 4. Minimize bycatch Seafood Watch defines bycatch as all fisheries-related mortality or injury other than the retained catch. Examples include discards, endangered or threatened species catch, pre-catch mortality and ghost fishing. All discards, including those released alive, are considered bycatch unless there is valid scientific evidence of high post-release survival and there is no documented evidence of negative impacts at the population level. The fishery optimizes the utilization of marine resources by minimizing post-harvest loss and by efficiently using marine resources as bait. All forms of bait used must not be derived from any unsustainable sources, whether of a marine or terrestrial origin. 5. Have no more than a negligible impact on any threatened, endangered or protected species The fishery avoids catch of any threatened, endangered or protected (ETP) species. If any ETP species are inadvertently caught, the fishery ensures and can demonstrate that it has no more than negligible impact on these populations and does not detrimentally affect the recovery rate of such species. 6. Th fishery is managed to sustain the long-term productivity of all affected species. Management should be appropriate for the inherent resilience of directly and indirectly affected marine life and should incorporate data sufficient to assess the affected species and manage fishing mortality to ensure little risk of depletion. Measures should be implemented and enforced to ensure that fishery mortality does not threaten the long-term productivity or ecological role of any directly or indirectly affected species in the future. The management strategy has high chance of preventing declines in stock productivity by taking into account the level of uncertainty, other impacts on the stock, and the potential for increased pressure in the future. The management strategy effectively prevents negative population impacts on bycatch species, particularly species of concern and significant impacts on ecosystem structure or function. 7. Avoid negative impacts on the structure, function or associated biota of marine habitats where fishing occurs. The fishery does not adversely affect the physical structure of the seafloor or associated biological communities. If high-impact gears (e.g. trawls, dredges) are used, vulnerable seafloor habitats (e.g. corals, seamounts) are not fished, and potential damage to the seafloor is mitigated through substantial spatial protection, gear modifications gear deployment and/or other highly effective methods and such mitigation is demonstrated. 8. Maintain the trophic role of all marine life All stocks are maintained at levels that allow them to fulfill their natural ecological role and to maintain a naturally functioning ecosystem and food web, as informed by the best available science and applied with appropriate levels of precaution. 9. Do not result in harmful ecological changes such as reduction of dependent predator populations, trophic cascades, or phase shifts Fishing activities must demonstrably not result in harmful changes such as depletion of dependent predators, trophic cascades, or phase shifts. This may require fishing certain species (e.g. forage species) well below maximum sustainable yield and maintaining populations of these species well above the biomass that produces maximum sustainable yield.

4 Version October 22, 2014

10. Ensure that any enhancement activities and fishing activities on enhanced stocks do not negatively affect the diversity, abundance or genetic integrity of wild stocks Any enhancement activities are conducted at levels that d not negatively affect wild stocks by reducing diversity, abundance or genetic integrity. Management of fisheries targeting enhanced stocks ensure that there are no negative impacts on the wild stocks, in line with the guiding principles described above, as a result of the fisheries. Enhancement activities do not negatively affect the ecosystem through density dependent competition or any other means, as informed by the best available science.

Seafood Watch Criteria for Fisheries

Public comment guidance - This section contains the standards preamble. We are not requesting specific feedback on this section, but feel free to comment on this general background information. Comment

The Monterey Bay Aquarium is committed to inspiring conservation of the oceans. To this end, Seafood Watch®, program of the Monterey Bay Aquarium, researches and evaluates the sustainability of fisheries products and shares these seafood recommendations with the public and other interested parties in several forms, including regionally specific Seafood Watch pocket guides, smartphone apps and online at www.seafoodwatch.org.

The criteria laid out in this document allow assessment of the relative sustainability of wild-capture fisheries according to the guiding principles and conservation ethic of the Monterey Bay Aquarium. Farmed seafood sources are evaluated with different set of criteria.

Seafood Watch® defines “sustainable seafood” as seafood from sources, whether fished or farmed, that can maintain or increase production without jeopardizing the structure and function of affected ecosystems. Sustainable wild-capture fisheries should ensure that the abundance of both targeted and incidentally caught species is maintained in the long term at levels that allow the species to fulfill its ecological role* while the natural structure, productivity, function and diversity of the habitat and ecosystem are all maintained. A management system should be in place that enforces all local, national and international laws to ensure long-term productivity of the resource and integrity of the ecosystem by adhering to the precautionary approach and responding to changing circumstances.

Scope Seafood Watch® recommendations apply to single stock or species caught in single fishery as defined by gear type, region and management body.

5 Version October 22, 2014

Table of Contents

Public comment guidance – The contents page provides an opportunity to see the full list of fisheries criteria. We welcome feedback on this list at the broad level of inclusion; i.e. are these the correct impact categories to be considering in an assessment of fisheries sustainability? Note that Seafood Watch does not currently intend to assess socioeconomic impacts and defers to other organizations with specific expertise, and that we are developing criteria to assess greenhouse gas (GHG) emissions associated with up-to-the-dock or farmgate production of seafood products from fisheries and aquaculture (available for public comment in a separate document). Comment agree with these broad categories of assessment. There is however significant opportunity to improve the way in which the sub-issues are distributed and represented in the Factors. My most significant observation about this would be about the way in which Factor 4.3 (at least how I interpret the meaning of ecosystem-based management in the fisheries context) is placed in the SFW assessment structure and the parameters/benchmarks applied. As it stands, this matter is ‘diluted’ by placing it in Criterion 4, whereas, to more fully implement the intent of Criterion 1 and 2 I consider that the practical implementation of EBM needs to be primarily included in these two latter Criteria. Much of my commentary and input below is directed towards explaining why I consider this important, and providing suggestions about how to possibly implement this change.

Criterion is focused on assessing the target species, from the perspective of the effects of fishing on the populations of the target species, and on the consequent cascading effects on ecologically related species and ecosystems of which the target species is component, throughout its various life-stages. This latter aspect is not well represented in the existing construction, but could be reasonably represented in the future by incorporating guidance from the Lenfest Task Force generalized to broadly apply to a range of ecological structure and functions affected (potentially) by fishing, including for species other than ‘forage’ species. I suggest the Lenfest work should be generalized and incorporated in a precautionary way into a revised Criterion 1.

consider the findings of the Lenfest analysis should be both generalized to apply more broadly across ecosystem issues as well as applied here in Criterion for stock productivity purposes because biomass parameters (with suitable thresholds) are powerful tool for the precautionary management of fisheries across broad range of fishing types, issues and data/information levels. Lenfest focus on the purpose of making such assessments principally for providing precautionary approach to managing the impacts of fishing o ecosystem structure and function in high latitude high-production systems and the so-called ‘forage’ species, but the same types of issues apply to range of dependant predators and prey in many other ecosystem types and impact situations. While these ecosystem impacts issues could also be assessed under other Criteria later in the SFW assessment system (under trophic etc issues), it would appear inconsistent to be using the same parameters for a different set of purposes (albeit at different grading thresholds). Eventually, different parameters, and different proxies, may be developed for these two major sets of purposes, but at present it seems most efficient to assess both productivity and ecosystem protection issues associated with fishing by using biomass parameters/thresholds in Criterion 1, with performance grades appropriately

6 Version October 22, 2014

(precautionarily) assigned to biomass thresholds for both productivity and ecosystem protection.

The rest of my comments here assume that the Lenfest recommendations are broadly accepted and appropriate parts are considered for incorporation into the next version of the SFW criteria within Criterion 1. Specifically, this should involve the development of biomass thresholds for stock size and for precautionary protection for dependent species of predators and prey, and a precautionary interpretation of the combination of these into the respective SFW grades for assessment of performance.

On socioeconomic issues, while I understand the difficulties, there is a substantial overlap between IUU fishing activities and social/humanitarian issues in the fishing sector, and I recommend that future iterations of the SFW should consider a specific Criterion targeting IUU activities, or a at least providing IUU with a stronger profile in the ME criterion. While this certainly wouldn't fully completely address the social/humanitarian issues associated with fishing, it would be a significant step in that direction, and the development process would provide a basis for subsequent further refinement.

I comment here throughout the text on a number of fine-scale matters that seem important, such as the vulnerability assessment based on the PSA approach, how to approach assignment of the burden of proof, and point to some possible issues with the lexicon. do not give each Criterion equal attention in this review, mainly because of a lack of time. For example, I haven’t been able to consider the bycatch criterion.

Given my suggestion for a major restructure of Criterion 1 based on biomass parameters/thresholds, and that this would affect the subsequent Criteria to some extent, it is hard for me to give an exact commentary, and some of my comments beyond Criterion 1 may not be completely relevant depending o what changes are subsequently made after the review. Nonetheless, I offer these comments in the spirit of constructive and incremental improvement, and I fully respect the role and responsibilities of the MBA team in sieving and incorporating aspects that seem useful. I also am aware and fully accept that there are many important aspects to be considered that go beyond the technical matters, and that strategic decisions always have to be made to assist with the overall efficiency and effectiveness of such an important performance assessment program as SFW.

Introduction...... 2 Providing feedback, comments and suggestion...... 2 Seafood Watch Ratings ...... 2 Seafood Watch Criteria for Fisheries ...... 5 Criterion 1 – Impacts on the Species Under Assessment...... 9 Factor 1.1 Inherent Vulnerability...... 10 Factor 1.2 Abundance...... 15 Factor 1.3 Fishing Mortality ...... 26 Criterion 2 – Impacts on Other Species...... 31 Factor 2.1 Inherent Vulnerability...... 34 Factor 2.2 Abundance...... 34 Factor 2.3 Fishing Mortality ...... 36

7 Version October 22, 2014

Factor 2.4 Modifying Factor: Discards and Bait Use ...... 38 Criterion 3 – Management Effectiveness...... 40 Factor 3.1 Harvest Strategy...... 42 Factor 3.2 Bycatch Management Strategy...... 49 Criterion 4 – Impacts on the Habitat and Ecosystem...... 53 Factor 4.1a Impact of Fishing Gear on the Habitat/Substrate...... 55 Factor 4.2 4.1b Modifying Factor: Mitigation of Gear Impacts...... 56 Factor 4.3 Ecosystem-based Fisheries Management...... 58 Overall Score and Final Recommendation ...... 63 Glossary...... 66 References...... 78 Appendices ...... 82 Appendix 1 – Further guidance on interpreting the health of stocks and fishing mortality ...... 82 Appendix 2 – Susceptibility attributes from the MSC FAM...... 85 Appendix 3 – Matrix of bycatch impacts by gear type...... 88 Appendix 4 – Appropriate management strategies ...... 98 Appendix 5 – Bycatch reduction approaches...... 104 Appendix 6 – Impact of fishing gear on the substrate ...... 109 Appendix 7 – Gear modification table for bottom tending gears ...... 113 Appendix References ...... 114

8 Criterion 1 – Impacts on the Species Under Assessment

Public Comment Guidance for Criterion 1

Overview – Criterion 1 is used for scoring the impacts of the fishery o the stock being assessed in the context of effects o the population of the target species It incorporates both the current abundance of the stock (i.e., whether it is overfished), and the fishing mortality (i.e., whether overfishing is occurring). The combination of scores for abundance and fishing mortality determines the score for Criterion 1. In addition, the same factors are assessed for all other major species caught in the fishery (including other targeted species as well as bycatch). These other species are assessed under Criterion 2, but the factors are identical. Therefore, the factors detailed below have particularly significant impact on the scoring of each fishery. These factors are also inherently complex as they take into account multiple considerations, including not just abundance and fishing mortality, but also inherent vulnerability of the species to fishing impacts uncertainty in the stock assessment or other data used to determine abundance and fishing mortality, and the degree to which fishery is one of the substantial contributors to the cumulative fishing mortality experienced by a species.

Because of its significance and complexity, many of the proposed changes for this criterion review focus o the structure and language in Criterion 1.

Specific proposals are available under each of Factor 1.1 and Factor 1.2. Please see these respective sections and provide comments there.

Feedback n specific proposals for Factors 1.2 and 1.3 should be provided in the boxes under these respective factors. Please provide any general comments or suggestions regarding the structure of Criterion 1 below. Comments 1. I agree with the proposal to have this Criterion consist of the Factors: abundance and fishing mortality (F).

In terms of decision structure, I would argue that the abundance Factor should be more heavily weighted than F, since (at least usually) abundance is an outcome parameter (in the decision structure sense) and has the capability to contribute much more robustly to answering the focal question at this point of the decision problem. Amongst other matters, assessment of abundance as an outcome parameter integrates all pressures and sources of mortality within single assessment grade. Whereas F is an estimate of one pressure on stock and its population, and in decision terms, is an estimate of an input parameter and likely to be weaker proxy for the outcome of the question – is this fishery detrimentally affecting the stock/population of the targeted species? Seeking greater accuracy and precision in both is obviously appropriate, but only if the relative weightings are correct in decision structure. Also, it is certainly critical to manage in fishery and management efforts to control must be very substantial, but this is a different problem from assessing the condition of a fished population based heavily o F, and so the grading needs to be appropriately weighted in reaching an overall conclusion about stock status. suggest that score for should contribute n more than 50% of weight for a score on abundance in making an assessment of the condition of fished population (ie abundance contributes 0.66 and contributes 0.33 of the final Criterion score). suspect this weighting might also be important to provide a perception of natural justice to fishers - if these two Factors are equally weighted, then a fishery that falls below the yellow bar in abundance but scores well above the yellow

*Underlined terms have been defined in the glossary. Ctrl + Click to view. bar o F may then be assigned an overall yellow grade even though abundance is low. For consumers, this could be perceived as biased outcome from Criterion 1 (inferring the fishery was in acceptable condition beause of an acceptable trend in stock increase, even though biomass was low or unknown). As with all the Criteria, the weighting of sub-scores needs to be carefully considered to avoid counter- intuitive assessment outcomes or weighting of assessment findings that are not consistent with the intent of the Criteria or SFW overall.

Guiding principles

The stock* is healthy and abundant. Abundance, size, sex, age and genetic structure should be maintained at levels that do not impair the long-term productivity of the stock or fulfillment of its role in the ecosystem and food web.

Fishing mortality does not threaten populations or impede the ecological role of any marine life. Fishing mortality should be appropriate given current abundance and inherent vulnerability to fishing while accounting for scientific uncertainty, management uncertainty, and non-fishery impacts such as habitat degradation.

Assessment instructions

Evaluate Factors 1.1–1.3 under Criterion 1 to score the stock for which you want a recommendation. Evaluate Factors 2.1–2.4 under Criterion 2 to score all other main species in the fishery, including both bycatch and retained species as well as any overfished depleted endangered, threatened or other species of concern that are regularly caught in the fishery.

Note that if wild stocks are assessed only in combination with hatchery-raised populations, the health of the wild stock cannot b considered better than “moderate concern”.

Factor 1.1 Inherent Vulnerability

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Public Comment Guidance

Overview Factor 1.1 is used to score the inherent vulnerability of species caught in the fishery, which is then used in Factor 1.2 to determine whether a species’ abundance is considered a moderate or high concern, when stock status is unknown We are proposing to integrate the current Factor 1.1 into Factor 1.2. Specific Proposed Changes Factor 1.1 is currently assessed for all species to determine if they are “high”, “medium”, or “low” vulnerability; however, this factor is only relevant to the scoring if the stock status of the species is unknown, and then only if it is of “high” vulnerability (i.e. “medium” and “low” vulnerability species are scored the same). This results in unnecessary work for the assessor and causes confusion in the review process. We are proposing to combine Factor 1.1 with Factor 1.2 so that vulnerability will be assessed only for those species with unknown stock status. In addition, we are proposing several changes to how vulnerability will be assessed.

The text in Factor 1.1 below has been marked with strikethrough to denote that it will be removed from this section (and moved to Factor 1.2). Details and rationale for these proposed changes can be found under Factor 1.2, as that is where the assessment instructions will now reside.

Feedback Please provide feedback below on the option of integrating Factor 1.1 with Factor 1.2. To provide feedback on the specific language used in this factor, please see Factor 1.2 with the information now resides. Questions: • Do you believe it is a good idea to integrate Factor 1.1 with Factor 1.2, and score inherent vulnerability only when it will affect the overall score (i.e. when stock abundance is unknown)? Why or why not?

In principle, support the incorporation of F1.1 into F1.2, and the use of inherent vulnerability. However, in doing this, the weighting, the level of precaution required, and the thresholds for the grading benchmarks need to be carefully balanced. Where stock abundance is unknown in detail (such as no appropriate stock assessment has been conducted), and the assessment here defaults to ‘vulnerability’, then considerable precaution needs to be applied to the scoring, or the scoring needs to reflect a precautionary system of vulnerability assessment that is tailored to the species under assessment and its specific fishing system. This can be done by both adjusting the parameters being assessed and by setting the performance benchmark at an appropriately high level. The way the criterion is now proposed to be constructed seems to infer that fishing is more likely to impact on species of high vulnerability than species of low vulnerability, but this is only likely to be true if management systems are the same (use same parameters/benchmarks/reference points etc) or ignore vulnerability.

However, would prefer to shift ‘vulnerability’ to Criterion 3, which seems more logical place to assess the effectiveness of risk assessment process. For example, at quick look, it seems that shrimp fishery that I am familiar with would be assigned as ‘low concern’ at Step 1b, but this would be very risky, given that shrimp trawling is targeted at aggregations and uses fishing techniques that are very effective at quickly reducing populations that would otherwise likely make important ecological contributions to the local food web. Making a risk assessment (the PSA approach) a default outcome

*Underlined terms have been defined in the glossary. Ctrl + Click to view. parameter here for abundance in Criterion 1 is highly dubious, and I doubt it can be considered a suitably accurate proxy for the assessment of abundance outcome unless there is a much more precautionary set of thresholds assigned than are evident in the tables.

strongly prefer the biomass threshold approach (as in Lenfest report) for use in assessing abundance, and widening the scope of fisheries that could be assessed for abundance directly, without resorting to vulnerability. The information gradient would need to be further developed into a benchmark gradient to match the SFW Black to Green gradient (which should be reasonably achievable). This could include such proxies as age class distribution, assessed by comparison of the time series age structure in the catch to natural population age structure or an appropriate surrogate benchmark. There are a number of similar empirical measures (some discussed in the appendices) that can assist with interpretation of the quality of ‘abundance’ and can be applied across many types of species, fisheries and knowledge- base situations.

This area is very complex, given the likely number of different practical scenarios. I haven’t had time to simulate this scoring system with any of the fisheries that I know well, but my first impression is that there is a big risk inferred by defaulting to primary vulnerability-based sieve to determine a grade for abundance (in the absence of a good stock assessment or a set of good empirical proxies). would argue that in such circumstances, the assessment would be more robust by requiring evidence of high levels of abundance through at least two empirical surrogates (in effect not defaulting solely to vulnerability (or F) for a Criterion 1 grade, even if there is a measure of precaution applied). There is considerable current research activity in the area of empirical surrogates, and this should be closely considered before finalizing this criterion.

Comments:

Goal: Ensure fishing mortality an other management measures are appropriate for the inherent vulnerability of the stock.

Instructions: Where available, Seafood Watch uses FishBase “vulnerability” scores to assign score for inherent vulnerability of the stock. The FishBase vulnerability score is derived from Cheung et al. (2005) and is found at www.fishbase.org o the species’ page.

FishBase vulnerability score Corresponding Seafood Watch inherent vulnerability category 0–35 Low inherent vulnerability 36–55 Moderate inherent vulnerability 56–100 High inherent vulnerability

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Some fish and all invertebrates d not have FishBase score. For these species, use the steps below (modified from the MSC Productivity Attributes from PSA Analysis used in Risk-Based Framework. Source (MSC 2009)).

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Step 1: Assign a score for each productivity attribute based o the table below. Scores will be 1, 2 or 3 for each attribute you select (if an attribute falls in between categories, intermediate scores (e.g. 2.5) may be used if well justified). Once all scores are selected, average the values to arrive at an overall vulnerability score (from 1–3). If any attribute is unknown or any value falls into an NA category (for moderate to high fecundity), that attribute is omitted from the average.

Resilience Fish Invertebrates attribute Score = Score = 2 Score = 3 Score = 1 Score = 2 Score = 3 1 Average age > 15 5–15 yrs yrs >1 yrs 5–15 yrs yrs at maturity yrs Average 25 10–25 yrs 1 yrs 2 yrs 10–25 yrs 1 yrs maximum age yrs Fecundity 100 N/A N/A 10 eggs/ yr NA NA eggs/ yr Average > 300 100–300 < 10 cm N/A N/A N/A maximum size cm cm Average size > 200 40-–200 < 4 cm N/A N/A N/A at maturity cm cm Reproductive Live Demersal Broadcast Live bearer Demersal egg Broadcast strategy bearer eg layer spawner layer or spawner brooder Trophic Level >3.25 2.75-3.25 <2.75 N/A N/A N/A Density N/A N/A N/A Depensatory No Compensatory dependence dynamics at depensatory or dynamics at low population compensatory low population sizes (Allee dynamics sizes effects) demonstrated demonstrated demonstrated or likely or likely or likely

Step 2 Assign a Seafood Watch inherent vulnerability category of high, medium or low based o the overall inherent vulnerability value calculated above (the average of all attribute scores):

Overall inherent vulnerability score from analysis Corresponding Seafood Finfish Invertebrates Watch inherent vulnerability category 2.44-3 2.46-3 Low inherent vulnerability 1.80-2.43 1.85-2.45 Moderate inherent vulnerability 1-1.79 1-1.84 High inherent vulnerability

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Factor 1.2 Abundance

Public comment guidance

Overview – Factor 1.2 scores the abundance of species caught in the fishery, including whether it is above or below limit and target reference points, if known, or classified as overfished, threatened or endangered. The score from Factor 1.2, combined with Factor 1.3 (which assesses fishing mortality), determines the score for Criterion 1 – Impact on the Species Under Assessment

Specific Proposed Changes We are proposing several changes to Factor 1.2 in order to 1) simplify and streamline its assessment, and 2) allow for consideration of circumstances that are not well captured by the current criteria. Specific proposed changes and their rationale are described below. The text changes proposed can be seen below in red (added text) and strikethrough (deleted text). Note: the text from Factor 1.1 has been moved to this section, but is not shown in red except where we are proposing changes to the text that was previously in Factor 1.2.

Inherent vulnerability – As described under Factor 1.1 in this document, we are proposing integrating the inherent vulnerability calculation (currently under Factor 1.1.) with Factor 1.2, as inherent vulnerability is only used in conjunction with scoring abundance (when abundance is unknown). This will help streamline assessments and avoid unnecessary work and confusion associated with assessing inherent vulnerability when (in most cases) it does not affect the score. In addition, we are proposing a change to how inherent vulnerability is calculated. Our current criteria have two methods of assessing inherent vulnerability – 1) using fishbase whenever a fishbase vulnerability score is available, and 2) using a productivity scoring table (based on Productivity- Susceptibility Analysis methodology) for invertebrates or any fish which do not have fishbase vulnerability score. However, in practice we have found that the fishbase score is often quite conservative compared both to the PSA score and to expert opinion. We conducted detailed analysis to compare various published PSA approaches (productivity attributes only; susceptibility is considered under 1.3) with expert opinion and based o the results, we are proposing a combination approach that uses the fishbase score as an initial filter (as fishbase score is always conservative according to our assessment) and then further scores those species that are high vulnerability according to fishbase, using the productivity attributes from the PSA methodology published in Field et al. 2010 to determine their final vulnerability score. This methodology was found in our assessment to have the closest correlation with species-specific expert input.

Reducing table from five to four classification tiers (removal of “very high concern” category) - Currently, Factor 1.2 has five classification tiers ranging from “very high concern” (score of 1/5) to “very low concern” (score of 5/5). Stocks classified as “overfished” are classified as “high concern”, while those that are endangered or threatened are classified as “very high concern”. In practice, we have found that this distinction is not always as clear as it needs to be to support the differential scoring. For example, some species (e.g. Pacific bluefin tuna) are very severely depleted (e.g. below 5% of virgin biomass) but are not listed as endangered or threatened, perhaps because the information on their status is relatively new, or for political reasons. Others may be listed as endangered but new data shows that their population is recovering, yet there has not been enough time and/or enough data for delisting. These species are still of concern, but not any higher concern than severely

*Underlined terms have been defined in the glossary. Ctrl + Click to view. depleted species that are not listed. Finally, differences in procedures for listing species in different states and nations, combined with the often poor fit of international classification systems (e.g. IUCN) with finfish species, further complicates the distinction. As a result, we propose combining these categories.

The change from to categories will streamline the assessments, but potential changes to scoring for overfished and endangered species needs to be considered. Scoring is always scaled linearly from 1 to 5, with the lowest category scored at a 1 (with the exception of critical concerns); therefore, the proposal adjusts the scoring accordingly. While this may appear to result in lower scores, we are proposing concomitant changes to the scoring of Factor 1.3, and the scoring must be considered together. To illuminate this complex issue, we have provided a separate comment section with guidance and examples specifically on scoring of Criterion 1, and ask for your feedback on the scoring there. In this section, please provide any comments on the proposal to group overfished and endangered/threatened species together in the “high concern” category.

Consideration of trend data for species below target biomass levels – In the current criteria, stocks that are above limit reference point but below target reference point are scored a “low concern”. This is to avoid overly penalizing fisheries that have rebuilding stocks, especially when rebuilding may take time but the fishery has already implemented strong management, and to account for the consideration that even in well managed fisheries, stocks will fluctuate around target reference points but should not go below limit reference points. However, we have found this structure is not appropriate for the situation in which a stock is below target reference point but showing a clear declining trend, approaching the limit reference point. In such cases, we feel higher caution is warranted. As such, we are proposing that stocks in this situation are classified as a “moderate concern”.

Classifying high vulnerability, data-poor species – In the current criteria, high vulnerability species are classified as high concern under 1.2 if stock status is completely unknown. If there are data demonstrating that the stock is not overfished, they are classified as a low concern. We have found that these criteria do not account for cases where there may be some data indicating the species is not depleted (and therefore it is not appropriate to score as a “high concern”), but the data are not sufficient (due to high uncertainty, lack of stock assessment, etc.) to conclude the stock is necessarily a “low concern”. We are proposing adding guidance to the criteria to allow for scoring as a “moderate concern” in these cases.

Feedback: please comment below o these proposed changes as well as any other comments o this factor (including comments on inherent vulnerability). Please see specific questions for public consultation below as well.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Comments on above proposed changes Inherent Vulnerability As above, don't agree that using vulnerability in cases where stock assessments are weak or don't exist is a good surrogate for Criterion 1 for SFW. Amongst my many concerns, in addition to the decision analysis problem outlined above, the process outlined to determine vulnerability is coarse, and is based almost solely on matters of MSY and sustained production. Matters of ecological impacts of target species fishing are very poorly estimated by the PSA and the related procedures discussed under this Criterion, and I would argue that ecological impacts (and particularly trophic impacts) will require much more stringent constraints on abundance and fishing mortality than are reflected in the production benchmarks used in Criterion 1 (and the same point is well made in the clarifying discussion documents, and strongly in Lenfest report, but have not been crafted here into the Criteria assessment system).

fully understand the point about availability and quality of information, but that is not a reason for applying weak logical framework simply to admit fishery into an assessment, and reach a grading decision. Such an approach is classical Type 3 error (determining an accurate and precise answer to the wrong question), and in my view, needs to be avoided. It seems that vulnerability will be used to create a filter, which I agree with, but the default (in the absence of evidence to the contrary) is that the target species should be assigned at least as ‘moderate concern’ in the absence of acceptable proxies for abundance. So, this would mean that those proxies could be used to estimate abundance, and vulnerability would not be considered. Moving vulnerability to Criterion 3 would assist avoid this problem.

Apart from this decision complexity, the main problem here is that the biomass threshold for Criterion 1 (BMSY) has been historically used mainly for production purposes, and there is a big conceptual shift required (and a shift in the benchmarks) to enable biomass thresholds to be used to assess abundance in the target species population for both production and for ecological impacts purposes. In practice, a precautionary biomass benchmark for a fishery in relation to ecological impacts is likely to be much higher (as in Lenfest report, about 80% Bunfished) rather than the level for achieving MSY (about 40% Bunfished). In practice this also means that any assessment system for ecological impacts based on biomass thresholds will, by default, be likely to also subsume achievement of MSY biomass benchmarks. As result, the MSY biomass benchmark may be best assigned to either Red or bottom of the Yellow grade, as a form of SFW entry level performance condition, with biomass above MSY assigned with incremental grades, peaking at 80% Bunfished, presumably meeting the Green threshold. Of course, achieving these grades would need to be modified according to uncertainty in the information base, and this could be developed using the approach developed in part in the Lenfest report.

If this was to be attempted, it would not be big shift in the assessment structure, but would require some careful re-organisation of the factors and benchmarks. However, without this, the ecological trophic issues associated with fishing the target species to MSY are buried within Criterion 1, and swamped by the high resolution data on stock assessments etc which are all directed to MSY (or its derivatives) as benchmark, providing weak ecological inferences. The irony of this is that many commercial fisheries operating to these types of outcomes would most likely be able to increase harvests above their current yield, and there is a high probability of increased unit profit even though the yields would be sub-optimal in terms of the MSY short-term production biomass.Applying the MSY approach within the real world fishery management systems in a lot of fisheries has not produced the expected outcomes over the long term – many such fisheries, applying the ‘best practice’ of the day, are suffering continuing decadal declines in yields.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. 4-5 points have no specific concern about the proposal to shift from 5 to 4 point grade, although given the structural complexity of the indicators and assigned benchmarks, and given my comments above/below, suspect it would be better to maintain the 5 point scale, even if the difficult parameters are simplified to allow this to more easily implemented by the assessors.

Target and Limit Reference Points Clearly, the SFW assessment should consider where a stock is positioned against the reference points, as well as the quality of the information that results in an assessment of that position. However, the benchmark levels of abundance and the form of those benchmarks set as the reference points is equally as important as the quality of evidence is used to determined how compliant a stock is to the reference point quantities.

In terms of the specific question asked, I agree. Stocks trending down, even though above limits, should be of moderate concern. But this does depend o the reference points being correctly set and the trend being assessed robustly. would like to see some additional specific guidance provided to assessors about how to determine if reference points are valid in the context of specific fishery types, and some examples of the type of empirical proxies that can be used to provide in-principle confirmation that stock abundances and trends are within the expected boundaries.

Classifying high vulnerability, data-poor species don't agree in-principle with the proposed change, consistent with my arguments above.

On the matter of using PSA: I don't have any specific concerns with PSA per se but it has to be very carefully applied as a form of risk assessment. The extant implementations of PSA (including Micheli et al) use equilibrium and deterministic models and MSY type benchmarks (eg Patrick et al 2009) to inform risk assignment, in the combination of P and S that are determined as specific levels of risk. Such assignments take no account of the spatial or temporal heterogeneity of the range of either the direct or indirect fishing impact receptors, the ecologically important stochastic patterns in these receptors, or the dynamics of the impact pressures themselves. There has never been any form of verification that risk assignments produced by PSA as it is implemented (eg in the MSC system) provide a reasonable representation of ecosystem structure and function impacts. Indeed, in the few cases where such matters have been examined (the direct and indirect effects of fishing o cascading dependent species), although PSA could not be used (resolution too coarse to be effective), it is clear that PSA assessment would not have identified high level of risk. Therefore greater resolution in PSA assessments (eg through higher resolution of input parameters) seems unlikely to provide greater support for PSA outcomes. Indeed, higher resolution inputs are likely to create a false sense of security, are very likely to be counter-productive, and create a high risk of Type 3 error in the assessment. The PSA is good example of risk assessment system, and it is important in assessing risks to fishing productivity, but remains a very experimental and weak approach to identifying and ranking risks to ecosystem structure and function. If PSA is to be used in SFW then it would seem best used as a risk identification process, not a tool for grading actual risk and assigning scores to fisheries in relation to the abundance parameters. It does not seem to me to be a defendable strategy for SFW to use PSA as a surrogate for abundance as parameter to assess ecosystem risks. support the graded and careful use of PSA for risk purposes in fishing productivity assessments, but extending this to ecosystem risks is a very big stretch, and likely to very substantially underestimate such risks.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. In short, the use of PSA in its various current forms (Micheli, Patrick, etc) seems inappropriate if the assessment system is assessing fishery for its impacts (direct and indirect) o ecosystem structure and function. The use of PSA as proposed to be implemented here will likely protect against extinction of the target species, but provides very little protection for retention of any significant level of natural ecosystem structure or function, even taking account of cumulative effects of fisheries (as proposed by Micheli et al 2014). Amongst the many other issues underlying the PSA approach, the benchmarks typically applied are supported by neither empirical research nor ecological theory. The logical constructs usually presented to defend the decision framework are heavily based in equilibrium and deterministic models, or simple statistical distributions, and do not account for dynamics in fishing impacts, or the stochastic heterogeneity in space and time of ecosystem receptors. For example, there are many circumstances where the PSA would assign a fishery that exclusively targets breeding aggregations as low risk, whereas within an ecological framework such a fishery would normally be identified as a high risk in the absence of verifiable evidence to the contrary (as in the shrimp example mentioned above).

If this is considered a controversial or outlying position, then suggest that SFW designers should consider the situation where ocean regions consist exclusively of target species that are all maintained at 2 to 30% Bunfished abundance (a likely outcome of managing mainly for MSY, or using only the combined low-resolution PSA parameters). In the absence of evidence to the contrary, and assigning the burden of proof to fishing rather than to the environment, more precautionary approach would be to assign as acceptable a target condition for fished species that is much higher than 40% Bunfished, and preferably in the region of 80% Bunfished. There are various forms of logical and empirical ecological argument (derived from operations research and decision analysis) as well as number of empirical research papers in various fishing situations that support this assertion, (some alluded to above, and in the appendices) but which I don't have space/time to present here.

The issues and questions about reference points (the in-principle questions about where to set the performance bar in the SFW grading scale) are moot. In the framework I discuss above, a fishery operating at BMSY might be acceptable as a minimum score (say 1) while achieving higher levels of sustained biomass (and respective levels of lower fishing mortality etc) could be awarded high grades, u to a high score for 80% Bunfished biomass or above (perhaps confirmed by census data or appropriate empirical surrogates). Attempting to resolve the differences in importance of target or limit biomass is overly optimistic application of resolution. Only very well known fisheries will have robust data o this, and even then the inferences for management purposes may be disputable, and erecting a grading system to separate these (at the level of quartiles) seems futile. In the absence of robust data o direct and indirect impacts of target species abundance range o ecosystem components, the main issue that should be addressed here is the current level of biomass relative to unfished biomass (the central biomass parameter also advocated by Lenfest report), and assessed irrespective of relativity to target or limit biomass constructed to service MSY. Although the latter may be used as strategy to achieve the former, it serves no valuable purpose to indicate that MSY provides robust concept for protecting ecosystems from the impacts of fishing.

understand the issues surrounding the use of Fishbase and PSA to establish estimates of vulnerability, but neither of these is directly suitable for ecologically-based estimation of the capacity/vulnerability of changes in the target species to drive cascading ecologically dependant impacts, which has to incorporate a range of other pressures beyond direct effects of fishing. There are other developing

*Underlined terms have been defined in the glossary. Ctrl + Click to view. empirical systems for estimating vulnerability which may be more suitable for direct use in SFW-style criteria, and these could be investigated for application here.

A pragmatic way to proceed here would be to review the benchmarks hierarchy and the glossary in detail so that the terms and definitions are consistent with the ecosystem-based management approach, rather than just fishery production. For example, under reference points, the definitions should be changed to also refer to goals for ecosystems structure and function, viz: “Fishing mortality reference points should be designed with the goal of ensuring that catch does not exceed sustainable yield and has very low likelihood of leading to depletion of the stock in the future.” Should read: “Fishing mortality reference points should be designed with the goal of ensuring that catch does not exceed sustainable yield, has very low likelihood of leading to depletion of the stock in the future, and, has very low likelihood of leading to detrimental impacts o the ecosystem structure or function of the ecosystems and dependant predators and prey in any habitats occupied by each life stage of the target species”.

Criterion Development and Expert Opinion Informing the criterion development process, then when the parameters have been reasonably agreed, populating the benchmarks using expert opinion is potentially useful. The outcome of expert consultation inferred here seems to be heavily weighted in favour of the ‘best practice’ fishing approach. Presumably, the decision input has been sought from fishery specialists, and so focus on ‘best practice’ fishing approaches/standards is expected. For example, using PSA is a biased towards fishing-friendly outcomes relative to ecosystem outcomes, as it is based o fishery production criteria and standards. What the Lenfest report best demonstrates is that the ‘best practice’ approach will not achieve an adequate level of protection of the structure and function of ecosystems where fishing is conducted, and higher standards are required.

In my experience (and I have conducted number of high-stakes elicitation of expert opinion in the private and public sector, including in the area of fisheries sustainability) gathering expert opinion on these matters can be very useful, but there are be a number of important pre-conditions that need to be well satisfied. The design of such consultations needs to be very carefully constructed, and if this is to be extended into the areas of EBM in fisheries, it would need to be very carefully managed to be technically defendable, and to avoid interest group capture through mission creep and decibel democracy. good recent model in my experience is: The rapid assessment workshop to elicit expert consensus to inform the development of the Great Barrier Reef Outlook Report 2014. Ward, 2014. Great Barrier Reef Marine Park Authority. http://elibrary.gbrmpa.gov.au/jspui/handle/11017/2858

Questions for public consultation

1. Do you think it is appropriate to score inherent vulnerability: a) only when stock status is unknown (as proposed here), b) for all assessed species, or c) not at all?

2. Please provide any thoughts on the proposal for scoring inherent vulnerability.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. 3. In some cases, we have used expert opinion to override the fishbase score where that score was deemed to be inaccurate for a given species. Even though using the fishbase score as a filter combined with the Field et al 201 productivity metric is expected to provide better correspondence to expert opinion, this method for scoring vulnerability may still disagree with expert opinion in some cases. In such cases, do you believe it is ever advisable to override the vulnerability score based on expert opinion? If so, in what circumstances would you seek expert opinion and use it as the basis of the ranking rather than using the method defined in the criteria?

Please refer to my commentary above for these questions above.

4. The NMFS methodology (currently under development/not yet finalized) uses combination of sensitivity attributes and climate factors as expert opinion to assess vulnerability of a fish stock to climate change, and is referred to as the “Fish Stock Climate Vulnerability Assessment”. A description of the methodology can be found here: http://www.st.nmfs.noaa.gov/ecosystems/climate/activities/assessing-vulnerability-of-fish- stocks.

1 page fact sheet can be found here: http://www.st.nmfs.noaa.gov/Assets/ecosystems/climate/documents/Fish_Stock_Climate_Vuln erability_Assessment.pdf

Do you feel this index, once developed, should be integrated into our assessment of inherent vulnerability? If so, how?

There are number of such approaches in development, but I haven’t had time to review or compare them in detail to the NOAA approach.

5. We are hoping to provide further clarification and guidance on what level and type of uncertainty should be considered evidence that a stock should be scored “low” rather than “very low”, or “moderate” rather than “low concern”. Considerations may include age of the stock assessment, availability of fishery independent vs fishery dependent (e.g. CPUE) data, scientific controversy, etc. Please provide any suggestions or feedback on these considerations, or additional considerations for measuring uncertainty?

The term ‘stock assessment’ is (in Australian parlance) very ‘freighted’. When a fishery is subjected to stock assessment, a very well controlled and directed process is invoked, and is always implemented by the technical (fishery science) experts. The usual approach is heavily steeped in MSY culture, and this is usually the target condition, often coded into surrogates and derivatives (eg in terms of spawning biomass) with very rare consideration (at best qualitative) of the direct and indirect dependencies. Very rarely is there any high-level ecological expertise involved. So, my response to this is: yes, use stock

*Underlined terms have been defined in the glossary. Ctrl + Click to view. assessments, but look critically at how they have been implemented, look carefully at all input parameters and their levels, look critically at assumptions and boundary conditions that are encoded into the models, and closely consider the purposes for which stocks are being assessed. Assessors need to be aware that most stock assessments do not address ecological structure and function benchmarks beyond high levels of extinction risk, only MSY production benchmarks. Age of the assessment seems a very minor point compared to these matters.

6. We have received mixed feedback from scientific advisors as to whether or not any full stock assessment should be considered as valid (i.e. low uncertainty), with “high uncertainty” reserved or situations that lack stock assessments but may have other information such as FAO status available. What is your opinion? If you do not believe all stock assessments are low uncertainty please share any thoughts on how to rate the quality of stock assessments based on considerations such as model type, validity of and sensitivity to model inputs, confidence intervals of model outputs, etc. All normal forms of stock assessment are ‘high uncertainty’ with respect to ecological consequences of fishing. The typical ‘best practice’ stock assessment should never be considered to provide ‘low uncertainty’ in relation to stock abundance, because of limitations on how externalities are treated and forecast. 7. Do you have any feedback on the proposal to eliminate the “very high concern” category? would probably prefer to see any such fishery not even mentioned, but this will mean making sure that the standard applied can identify such fisheries, and that the assessment system is capable of simply dropping them from an assessment so that resources are not wasted.

8. Do you have any feedback on the proposal to use trend data to distinguish between stocks that are rebuilding, versus those that are below target levels and declining? Trending should not be included as a specific assessment parameter, unless grades are assigned with low weighting. Trends downwards is an important signal, but this (and the inverse) should only be recognized when abundance is above an acceptable (probably MSY) threshold. The trends currently shown by a fishery may not persist, and by grading ‘trending’ fishery that is currently below satisfactory abundance standards, fisheries that are already above the standard will perceive disadvantage, and its not clear to me how to convince consumers of the relative value of ‘incremental progress’ if abundance is low. Perhaps a separate category for ‘improving’ might be erected and applied as modifying factor to abundance grades, but not to independently infer acceptability. am not convinced that there is any trend that could be included in a fishery assessment as being an acceptable contribution to the fisheries status, in terms of equity in the decision problem. Of course, a fishery trending upwards in biomass is good, and one trending down is not good, but the main question is still – at what level is the biomass? trending index that involves proportion of benchmark achievement might be feasible (trending down at 90% of the abundance benchmark would be of lesser concern that the same slope of decline 50% the benchmark), but I don't know of any precedent for this.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Irrespective, the specific issue of relevance might also be considered as one of management effectiveness – how does the management respond to signals of declining abundance, and at what place in the abundance patterns? So while it might be conceivable to develop a trending grade based on slope and current level relative to Bunfished, any substantive weighting of such score will possibly disadvantage a fishery that achieves a similar score based o abundance alone. Overall, am not convinced that any form of credit should be applied to stocks that are ‘rebuilding’ but remain below MSY, as, in my assessment experience, timelines and evidence are often vague, and there is considerable advantage to be secured by management of the information base to avoid any inference of negative achievement.

9. If you agree with using trend data as described above, do you have any suggested guidance for determining a declining trend (e.g., time period, whether decline must be monotonic, what method to use for determining if it is statistically significant, threshold rate of decline (steepness or slope) etc.?)

10. Do you believe that it would be appropriate to score highly vulnerable species with weak evidence that they are not overfished as “moderate concern”, rather than as “low concern” (as lower vulnerability species would score) or “high concern” (as high vulnerability species currently are scored in cases where there are no clear data on abundance relative to reference points)?

In my view ‘highly vulnerable species with weak evidence that they are not overfished’ are of ‘high concern’, because the benchmarks likely to have been used to assess overfishing combined with the weak evidence base is likely represent a high threat to ecosystems in the absence of evidence to the contrary.

Any other comments

Step 1: If species is of unknown abundance, assess its vulnerability. If species is of known abundance, vulnerability does not need to be assessed, and you can continue to Step 2.

1a: If FishBase “vulnerability” score is available, use this score to continue as follows: Note: The FishBase vulnerability score is derived from Cheung et al. (2005) and is found at www.fishbase.org o the species’ page.

FishBase Corresponding Seafood Watch Action vulnerability score inherent vulnerability category

*Underlined terms have been defined in the glossary. Ctrl + Click to view. 0–55 Low to moderate inherent Continue to Step 2. vulnerability 56–100 Inherent vulnerability may be Continue to Step 1b. high No fishbase score Inherent vulnerability may be Continue to Step 1b. high

1b: For species that have “high” vulnerability classification under fishbase, or no fishbase score, continue with the PSA approach below (adopted from Field et al. 2010) to classify its vulnerability.

Assign a score for each productivity attribute based o the table below. Scores will be 1, 2 or 3 for each attribute you select (if an attribute falls in between categories, intermediate scores (e.g. 2.5) may be used if well justified). Once all scores are selected, average the values to arrive at an overall vulnerability score (from 1–3). If any attribute is unknown or any value falls into an NA category (for moderate to high fecundity), that attribute is omitted from the average.

Resilience Fish Invertebrates attribute Score Score = 2 Score = 3 Score = 1 Score = 2 Score = 3 1 Average age 4 yrs 2-4 yrs 2 yrs >1 yrs 5–15 yrs yrs at maturity Average 40 20-40 yrs 20 yrs 2 yrs 10–25 yrs 1 yrs maximum yrs age Fecundity 100 N/A N/A 10 eggs/ N/A N/A eggs/ yr yr Average 80 40-80 cm 40 cm N/A N/A N/A maximum cm size Reproductiv Live Demersal Broadcast Live bearer Demersal egg Broadcast strategy bear- eg layer spawner layer or spawner er brooder Trophic Level >3.5 2.5-3.5 <2.5 N/A N/A N/A Von 0.10- N/A N/A N/A 0.10 > 0.20 Bertalanffy 0.20 Growth (K) Natural 0.10- N/A N/A N/A 0.10 > 0.20 Mortality 0.20 Rate (M) Density N/A N/A N/A Depen- No Compen- dependence satory depensatory satory dynamics at or compen- dynamics at low satory low population dynamics population 24

*Underlined terms have been defined in the glossary. Ctrl + Click to view. sizes (Allee demon- sizes demon- effects) strated or strated or demon- likely likely strated or likely

Assign a Seafood Watch inherent vulnerability category of high, medium or low based o the overall inherent vulnerability value calculated above (the average of all attribute scores):

Step 2: Score according to chart below. If species is of unknown abundance, use the vulnerability category calculated above.

Conservation Description Score Concern Very Low There is reliable quantitative stock assessment, and biomass is estimated to 5 be above or fluctuating around an appropriate target reference point with no low uncertainty scientific controversy around that estimate, or Stock is at or very near its historic high or virgin biomass estimated with levels of uncertainty appropriate to the type of fishery and species under assessment. or Species is non-native Low Stock is classified as not overfished but quantitative stock assessment is • 4 lacking, • 3.67 or Biomass is above the limit reference point but may be below target reference point, and there is clear increasing trend over >1 generation time, all estimated with levels of uncertainty appropriate to the type of fishery and species under assessment.

or Biomass is securely above the limit reference point and may be estimated to be above a target reference point but there is significant uncertainty (e.g., widely varied results depending on model or assumptions) or scientific controversy around that estimate or around the suitability of the target reference point, and species is not highly vulnerable 25

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Moderate There is no evidence to suggest that stock is either above or below reference 3 points; Unknown and Stock inherent vulnerability is not high moderate or low 2.33 (as scored in Factor 1.1).

Species is above limit reference point but below a target reference point, and there is a clear declining trend or n increase over >1 generation time,[this is consistent with my amendment to ‘Low’]

There is conflicting information about stock status (e.g., IUCN listing disagrees with stock assessment), and there is no clear reason to suggest one set of information is more valid

Species is highly vulnerable AND stock status is not known, or is highly uncertain (i.e. there is no quantitative, robust stock assessment) but there are some data indicating status is not of concern (e.g. increasing population trends, IUCN least concern status, or a stock assessment indicating stock is above a limit reference point, but with high uncertainty) High Probable that stock is below the limit reference point, 2 1 or Stock is listed by management body as overfished or depleted, or is stock or species of concern (as listed by state or federal management body) an IUCN Near Threatened or Vulnerable species, or equivalent, or There is no evidence to suggest that stock is either above or below reference points; Unknown and Stock inherent vulnerability is high (as scored in Factor 1.1). OR Stock or species is determined by a state, national or international scientific body to be endangered or threatened. Very High Stock or species is listed by a state, national or international scientific body as 1 endangered or threatened.

Factor 1.3 Fishing Mortality

Public Comments Guidance

Overview – Factor 1.3 assessing whether fishing mortality for each species caught in the fishery is at an appropriate level (defined as at or below maximum sustainable yield, or an equivalent proxy). 26

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Because these same criteria are used to assess not only targeted species, but also bycatch species (under Criterion 2.3), it is necessary to provide guidance even for species that may be a relatively minor component of the catch and also are impacted by other fisheries. We distinguish between whether or not the fishery under assessment is a “substantial contributor” (which includes all target fisheries, as well as those that are one of the main contributors to fishing mortality experienced by a species; see glossary definition).

In addition to providing separate tables for assessing species depending on whether or not the fishery is “substantial contributor” to mortality, the current criteria also integrate information on whether the stock is depleted, and management is in place. One goal of this proposal is to streamline this factor by focusing solely on fishing mortality. In addition, we are proposing to combine the “low” and “very low” concern in order to further simplification assessment of this factor.

Specific Proposed Changes: 1. Simplification of the fishing mortality table – Depleted status and whether management is in place when overfishing is occurring are accounted for in other criteria (1.2 and 3.1 respectively). Including those criteria here double-counts these issues and makes this factor unnecessarily complex. We have removed the language regarding these issues in this proposal below. We have checked that these considerations are adequately addressed in factors 1.2 and 3.1. For example, fishery that has catch of depleted species, experiencing overfishing, without sufficient management will get a “poor” in management criterion 3.1 which functions as a “critical” (i.e. determines an overall Avoid ranking); therefore, the separate critical score here is not needed. See the separate comment box o Scoring of Criterion 1 for examples of how these fisheries would score under the current and proposed new criteria. Combining low and very low concern categories – Currently, the probability that overfishing is not occurring is used to distinguish “very low concern” from “low concern”. In practice, this information is often not available and it can be difficult to distinguish between “low” and “very low” concern. When they are distinguished, “low concern” species often end up scored too harshly. These species are still considered to be at level where fishing mortality isn’t threatening the population, yet because they receive a decreased score for Factor 1.3, they may receive a low yellow or even a red score when the species is endangered or overfished (this is primarily a concern for fisheries that have a low, but not negligible, bycatch of overfished or endangered species, which is not adversely impacting that species but still greatly lowers the overall score for the fishery). Combining the categories so that all species with fishing mortality

at or below sustainable level (i.e. FMSY or equivalent) get full credit would help rectify this scoring concern as well as simplify assessments. See the separate comment box on Scoring of Criterion 1 for examples of how fisheries would score under the current and proposed new criteria. Guidance for alternative reference points – Our current criteria are based on MSY-based reference points, and there is guidance in the appendices that reference points that are not as

conservative as FMSY should not be considered equivalent. However, we felt this guidance needed to be more clearly stated in the criteria table itself to provide more consistency in the common situation that other reference points are used. We have proposed adding language to

*Underlined terms have been defined in the glossary. Ctrl + Click to view. the “moderate concern” category to account for the case where F is below a reference point,

but that reference point is known to be less precautionary than FMSY In this case, whether fishing mortality is at a sustainable level must be considered unknown.

Feedback – please provide comments o the proposed changes, and any other comments o this factor, in the box below. Questions for Public Comment 1. Please provide any thoughts on the simplification of Factor 1.3 by removing references to the abundance status and management of species:

2. Please provide any feedback on the proposal to combine the “low concern” and “very low concern” categories:

Any Other Comments As described above, this criterion needs to be substantially revised, or the intent statements for the SFW program need to substantively changed. It seems unwise to assess fishing mortality and weight it equivalently to abundance, or replace an abundance assessment with fishing mortality grade in the absence of supporting confirmatory evidence.

Goal: Fishing mortality is appropriate for current state of the stock.

NOTE: Ratings are based o fishing mortality/exploitation rate, e.g., F/FMSY For the purposes of this table, a stock is deemed “not depleted” if rated as Very Low to Moderate Conservation Concern under 1.2; an is deemed “depleted” if rated as High to Very High Conservation Concern under 1.2. When determining whether a fishery is a substantial contributor and/or whether fishing mortality is at or below a sustainable level, err o the side of caution when there is uncertainty. For further guidance, see Appendix 1 (guidance on evaluating fishing mortality).

For Target Species or When Fishery is substantial contributor to Mortality (include cases where fishery is one of many fisheries that equally contribute to overall mortality) Conservation Description Score Concern Very Low • Highly likely that fishing mortality is at or below a sustainable level that will 5 Concern allow population to maintain current level or rebuild if depleted, OR • Large proportion of population is protected [NOTE: large proportion protected suffices only if not depleted], OR

*Underlined terms have been defined in the glossary. Ctrl + Click to view. • Species is non-native Low Concern • Probable (>50% chance) that fishing mortality is at or below a sustainable 3.67 level that will allow population to maintain current level or rebuild if 5 depleted, but some uncertainty, OR • Population trends are increasing in short and long term due to management OR • Large proportion of population is protected. [NOTE: large proportion protected suffices only if not depleted], OR • Species is non-native Moderate • is fluctuating around FMSY, OR 2.33 Concern • is unknown, but for any depleted populations, effective management is in 3 place [Note: If F is unknown and population is depleted, effective management must be in place to get ‘moderate’ score]

• is below reference point but reference point is less conservative than FMSY

High Concern • Overfishing is occurring/cumulative fishing pressure may be too high to 1 allow species to maintain abundance or recover, but for any depleted populations, management that is reasonably expected to curtail overfishing is in place (make sure does not meet critical rating below), OR • is unknown, population is depleted, and no effective management in place Critical Overfishing is occurring/cumulative fishing pressure may be too high to allow 0 species to maintain abundance or recover; population is depleted; and reasonable management to reduce fishing mortality/curtail overfishing is not in place.

For non-target species in which fishery contribution to mortality may be low or unknown Conservation Description Score Concern Very Low • Fishery’s contribution to mortality is negligible, OR 5 Concern • Meets definition of “Very Low Concern” in table above Low Concern • Fishery contribution to mortality may not be is negligible or but does not 3.67 adversely affect population, OR 5 • and/or Fishery contribution is unknown, but population is not depleted and susceptibility (see Appendix 2 to fishery is low, OR • Meets “Low concern” in table above. Moderate Fishery contribution F is unknown, but population may be depleted (and if so, 2.33 Concern management is in place) or susceptibility to fishery is moderate to high (see 3 Appendix 2) This level of performance seems too low for ‘moderate concern’. Irrespective of estimates of susceptibility (which consider to have very wide bounds of confidence and relevance), any fishery where F is unknown should be considered a high risk, and without evidence to the contrary, of high concern High Concern Cumulative overfishing is occurring, fishery contribution is unknown, and 1 susceptibility to fishery is moderate to high (see Appendix 2) but population is depleted and n reasonable management to curtail overfishing is in place.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Criterion 1 Score and Rating

Public Comment Guidance – Scoring of Criterion 1.

Both Factor 1.2 and Factor 1.3 contain proposals to simplify assessments by reducing the number of scoring categories and adjusting the numerical scores (which are scaled linearly from 1 to 5) accordingly. Because scores are assigned for Criterion 1 based o the geometric mean of Factor 1.2 and Factor 1.3, these proposed changes would interact to affect the score and color (red, yellow or green) rating for Criterion 1. Below we have included some examples of how hypothetical fisheries would score under the current criteria as well as with the proposed changes.

Description of hypothetical fishery Current Criteria Rating Proposed Criteria (Score) Rating (Score) Bycatch species that is endangered, and is caught Red (1.92) Yellow (2.24) infrequently or at a level that does not adversely affect the species (but catch is not “negligible”) Overfished species with overfishing occurring Red (1.41)/Critical (0) Red (1) depending on management Stock status and fishing mortality are unknown Yellow (2.63) Yellow (2.64) Species that is below target but above limit Yellow (3.05) Green (3.29) reference point and increasing, fishing mortality is unknown Species that is below target but above limit Yellow (3.05) Yellow (2.63) reference point and decreasing, fishing mortality is unknown

Feedback: Please provide any feedback o these scoring examples in the box below. Comments: Given that I have argued for a major reconstruction of this Criterion, I cannot rationally comment on the standard that is encoded here.

My main point remains that F is an input to the decision analysis problem, and should be down- weighted with respect to abundance, which is an outcome. No amount of increased accuracy and precision in F should substitute for an assessment of abundance, and even a weak set of estimates for the abundance based o proxies are better than accurate estimates of F. Even if was known without error, translating this into an estimate of the full range of ecological impacts of fishing the stock is highly dubious, and defaulting to F is not a defendable basis for the SFW assessment of abundance, given that this is intended to be a proxy for population and ecological impacts of fishing on the target species. None of this is to argue that F is unimportant – clearly F is the central tool of fisheries management, and should be a major parameter of the ME Criterion.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Score geometric mean (Factors 1.2, 1.3). Factor 1.1 is not included in the calculation; instead it modifies the score of Factor 1.2 in some circumstances (See Factor 1.2 table for more information).

Rating is based o the Score as follows: • >3.2 Green • >2.2 and <=3.2 Yellow • <=2.2 Red

Rating is Critical if Factor 1.3 is Critical. Criterion 2 – Impacts on Other Species

Guiding principles

The fishery minimizes bycatch. Seafood Watch® defines bycatch as all fisheries-related mortality or injury other than the retained catch. Examples include discards, endangered or threatened species catch, pre-catch mortality and ghost fishing. All discards, including those released alive, are considered bycatch unless there is valid scientific evidence of high post-release survival and there is no documented evidence of negative impacts at the population level.

Fishing mortality does not threaten populations or impede the ecological role of any marine life. Fishing mortality should be appropriate given each impacted (bycatch) species’ abundance and productivity, accounting for scientific uncertainty, management uncertainty and non-fishery impacts such as habitat degradation.

Assessment instructions

Public Comment Guidance - Criterion 2 Instructions

Overview – Criterion 2 is assessed using the same criteria as Criterion 1, but is applied to bycatch and other main species that are caught together in the fishery with the species under assessment (but see specific guidelines for marine mammals, and for cases where bycatch species are unknown). The most challenging aspect of assessing Criterion is determining which species need to be included for assessment. All “main species” are currently required to be fully assessed. This can be a very intensive effort, and the score for Criterion 2 is determined solely by the “worst case” species.

Specific Proposals – Though we do not have specific proposals for defining “main species” yet, we intend that a focus of the criteria review will be revisiting the definition of “main species” to determine a better filter for which species need to be assessed (specifically, one that is easier to assess even in data- limited situations), and one which would streamline assessments by helping analysts to assess only the key species that are likely to limit the score for Criterion 2 (without running the risk of overlooking a potential limiting species).

Feedback – at this time we are seeking all comments and suggestions for how to fine-tune this approach. Please provide comments below.

Questions: Please provide suggested changes to the definition of “main species” (see current definition below): 31

*Underlined terms have been defined in the glossary. Ctrl + Click to view. consider that the biomass basis for identifying ‘main’ species is one of two important aspects of the issue to be addressed - the other part is the ‘importance of impacts’ basis. This latter one is of course hard to estimate, but there are a number of well known parameters that could be erected as proxies: propensity for periodic/spatial aggregations for breeding, feeding; predator diet dependency profiles (as also deployed in the Lenfest report); etc. Even though species may only be taken in small numbers or biomass (relative to the target species) this does not mean that the fishery has only insignificant impacts o that species. As a result, a precautionary approach is needed here in identifying ‘main’ species, and the proportion biomass rules should be extended to include species that may have unique features that make them susceptible to either direct or indirect trophic effects of being part of the bycatch. As used in other systems, discriminating potential species impacts based o the so-called ‘species-overlap’ is notoriously weak, based o equilibrium and deterministic models of impacts, and provides only the weakest possible inference about the significance of ecological risk. To better inform the species to be considered, ‘second-tier’ group of species from the bycatch might be useful, based on attributes such as species that are known to have periodic aggregations that may overlap the target species, species that have specific and constrained habitat requirements a particular points in their lifecycle, that naturally occur only in low densities, that forage intensively in any area also targeted by the fishery, or species that occur in more than 1%, 5%, 10% etc of individual fishing operations. The performance grade could (for example) be constructed o these parameters to reflect the number of such second-tier species. This concept obviously needs further development, and could build on the vulnerability lists presented below, but could be an initial entry point for expanding the assessment of bycatch in a more ecologically relevant way.

Please provide other suggestions for how to determine which species need to be assessed (i.e. suggest an alternative approach to defining “main species”, assessing each of them and basing the score on the lowest scoring species):

Other comments: It seems to me that this Criterion could be re-named ‘Impacts on By-catch’ (or ‘…other Capture Species’), supported by the catch-all definition that is applied to include all forms of mortality in any species caught other than the target species. recognize that there are many interpretations of these terms, but an appropriate entry in the Glossary would be adequate to set the decision/assessment boundaries.

The Criterion score for the stock for which you want recommendation is the lowest score of all the other main species caught with it, multiplied by the discard rate. species is included in the assessment as main species if: • The catch of the species in the fishery under assessment composes >5% of that fishery’s catch, or • The species is >1% of that fishery’s catch and the fishery causes >5% of the species’ total mortality across all fisheries, or • The species is <1% of that fishery’s catch and the fishery causes >20% of species’ total mortality across all fisheries, or • The species is overfished, depleted, stock of concern, endangered, threatened, IUCN Near Threatened, US MMPA strategic species, and/or subject to overfishing and the fishery causes >1% of species’ total mortality across all fisheries.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. • If there are no other “main species” (based on the above guidance) besides the one assessed under criterion 1, but the total catch of other discarded and retained species is >5% (i.e. catch of criterion species is <95% of total), assess the top species by volume of catch (if there are only 1-2 other species caught, assess those species).

In some cases, bycatch or retained species in the fishery are not known. If species are unknown, Appendix 3 should be used to aid scoring. If some bycatch or retained species are known but information is incomplete (e.g., some or all retained species are recorded but bycatch species are not), assess each known species following guidance for “known species” and assess the fishery following guidance for “unknown species” under each Factor. Scoring for this criterion is based on the worst case, so the lowest of these scores will be used.

If there is no bycatch and no other species landed, the fishery receives a score of five for this criterion, the remaining questions in Criterion 2 can be skipped, and the assessor can continue with Criterion 3.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Factor 2.1 Inherent Vulnerability

Public Comment Guidance: This factor is based on Factor 1.1, with additional guidance provided for cases where there is bycatch of unknown species. As we are proposing to integrate Factor 1.1 into Factor 1.2, the same proposal applies here (to integrate Factor 2.1 into Factor 2.2).

Feedback – please provide any comments specific to Factor 2.1 (based o text below) here. If you have comments on the suggestions for Factor 1.1 or for integrating Factor 1.1/2.1 into Factor 2.1/2.2 as described in the Factor 1.1 section, please comment under Factor 1.1 Comments comment extensively on this above, and the same largely apply here.

Goal: Ensure fishing mortality an other management measures are appropriate for the inherent vulnerability of all bycatch stock(s).

Known species

Follow the assessment for Factor 1.1 above (the Factors for Inherent Vulnerability, Stock Status and Fishing Mortality are identical for all main species caught in the fishery, whether target, other retained or discarded).

Unknown species

Where the catch composition is not completely known, use Appendix 3 to identify those taxa that are most likely to interact substantially with the fishing gear, defined as taxa scoring a 3.5 or below. The numbers in the unknown bycatch matrix will be used extensively for scoring Criterion 2.3.

Vulnerability is assigned to the taxa that are identified as interacting with the fishery as follows:

High marine mammals, turtles, sharks, seabirds, deepwater and shallow biogenic habitat (seagrass beds, coral, sponges, etc.). Moderate invertebrates and fish.

Factor 2.2 Abundance

Public Comment Guidance: This factor is based on Factor 1.2, with additional guidance for cases where there is bycatch of unknown species.

Feedback – please provide any comments specific to Factor 2.2 (based o text below) here. If you have comments on the suggestions for Factor 1.2, please comment under Factor 1.2. Comments Same comments as earlier apply here.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Goal: Stock abundance and size structure of all main bycatch species/stocks is maintained at a level that does not impair recruitment or productivity.

Known species

Follow the assessment for Factor 1.2 above (the Factors for Inherent Vulnerability, Stock Status and Fishing Mortality are identical for all main species caught in the fishery, whether target, other retained or discarded).

Unknown species

If bycatch species and/or other retained species are unknown use Appendix 3, the unknown bycatch matrix, and additional bycatch guidance (in Appendix 3) to determine the types of taxa that are likely to interact with the fishery based on gear type and location. Factor 2.2 is scored as “moderate concern” for each taxon listed in Appendix 3 for this type of fishery with a score of 3.5 or below unless the taxon is comprised largely of species that are either of high vulnerability as scored in Factor 2.1, or overfished, endangered or threatened within the range of the fishery (e.g., sea turtles, seabirds, marine mammals and sharks); in these cases, it is scored as “high concern.” Species scoring 4 or above d not need to be addressed.

This is an important concept – establishing default set of bycatch species, and a good implementation of the burden of proof. In some ways it might be better to simply apply this generic list by fishery type and location (and improve it by extracting data from of all fisheries that have been assessed), and permit increased scores/grades if evidence is presented that such bycatch is not taken, or is taken but is not ecologically affected by the actions of the fishery.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Factor 2.3 Fishing Mortality

Public Comment Guidance for Factor 2.3

Overview – Generally, Criterion 2.3 follows the structure of Criterion 1.3, and the same changes proposed there will be applicable here as well. In addition, there is special guidance for marine mammals which have quantitative information and classifications when caught in US fisheries (managed under the Marine Mammal Protection Act). We are proposing one minor change to this special guidance.

Specific Proposal – SFW has determined that the guidance here is overly harsh in cases where total mortality for a species does not exceed its Potential Biological Removal (PBR). In these cases, fishery may be scored “moderate” in some instances; however, in all cases if PBR is not exceeded by cumulative fishing mortality, then the mortality is at a sustainable level, and thus a “low concern” would be most appropriate. Here we propose an edit such that all species that are experiencing cumulative fishing mortality levels below PBR are scored as a “low concern” for Factor 2.3 for all fisheries that interact with them. We also propose removing references to the “very low concern” category, as we are proposing removing that category from 2.3 altogether.

Feedback – Please provide comments on this proposed change below: Question Please provide any comments on the proposal to refine the guidance for marine mammals, as described above and shown in redlining below:

Any Other Comments o 2.3: PBR is concept heavily driven by productivity, and although it is an important tool for recovery of depleted species, it should not be used to infer an outcome to be achieved (in terms eg of population size, structure, distribution etc). This can be a risk for assessors, in terms of how to interpret management activities based on PBR.

Goal: Fishing mortality is appropriate for the current state of all main bycatch species/stocks.

Known species

Follow the assessment for Factor 1.3 above (the Factors for Inherent Vulnerability, Stock Status and Fishing Mortality are identical for all main species caught in the fishery, whether target, other retained or discarded).

Additional guidance for scoring of marine mammals caught in US fisheries is given below (due to the availability of data on potential biological removal (PBR) and fishing mortality rates on all bycaught

*Underlined terms have been defined in the glossary. Ctrl + Click to view. marine mammals, available in marine mammal stock assessments and List of Fisheries reports, see http://www.nmfs.noaa.gov/pr/interactions/lof/)

% of PBR taken by fishery Stock is “strategic” due to Seafood Watch Rating Fishery category in “List cumulative fisheries of Fisheries” (based on mortality >100% PBR? that species) <1% No Very Low Category III 1-10% No Very Low Cat II (if cumulative take >10%) or III (if cumulative take <=10%) 10-50% No Low Cat II 50-100% No Moderate Low Cat I <5% Yes Very Low Cat III 5-10% and not one of the Yes Low Cat II main contributors to total mortality 10-50% and not one of the Yes Moderate Cat II main contributors to total mortality >50% OR a main Yes High OR Critical depending Cat I contributor to total on management fisheries-related mortality If PBR or fishery mortality relative to PBR is not known, score conservatively given what is known (e.g., fishery and/or species classification) or score as “moderate”. Example: if it is unknown but fishery is classified as Category II and species is not strategic, score as “low”.

Unknown species

If bycatch and/or other retained species are unknown use Appendix 3, the unknown bycatch matrix, and additional bycatch guidance (in Appendix 3) to determine the types of taxa that are likely to interact with the fishery based on gear type and location. Each score from Appendix 3 has a corresponding fishing mortality conservation concern (below).

Appendix 3 the unknown bycatch matrix and additional guidance for unknown bycatch species should ONLY be used as a general guide to help rate bycatch potential (for Factor 2.3, mortality caused by the fishery) when no data on the composition of bycatch are available. In these cases, Appendix 3 provides guidance based on gear-type and its potential to interact with species other than the target fishery, as reported in the scientific literature. When assigning a conservation concern for unknown bycatch, other factors to consider before using the unknown bycatch matrix include: geographic range of the fishery, degree of overlap, if any (with foraging areas, breeding grounds, etc.) there is between fishing and potential bycatch species, fishing depth, whether the fishery is operating in coastal (some coastal areas may have a greater impacts on some species) or open-ocean systems, whether the fishery operates seasonally and coincides with breeding season, and other concerns based o fishing region and the conservation concern for the potential bycatch species. Species with scores above 3.5 from the unknown bycatch matrix d not need to be addressed.

Bycatch score from Moderate Concern High Concern species status (under 2.2) 37

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Appendix 3 (1–5) species status (under 2.2) 3-3.5 Low Concern Low Concern 2-2.5 Moderate Concern Moderate Concern 1 High Concern • High concern, if there are some efforts, believed to be effective, to reduce mortality • Critical, if there are no effective efforts to reduce mortality

Factor 2.4 Modifying Factor: Discards an Bait Use

Public Comment Guidance: Another goal of the criteria review, with respect to Criterion 2, will be to collaborate with the aquaculture team regarding the scoring for discard rates and bait use (Factor 2.4) to ensure that there is consistency with the scoring of feed use in aquaculture reports. In addition, because bait use is considered in 2.4 but is rarely quantified, we aim to provide default scores for bait use, based o literature review, for a variety of fishery types (target species and gear). We will provide an opportunity to override these default scores if data specific to the fishery can be provided.

Please provide any comments on these ideas or the current text of Factor 2.4 below.

Comments This is an important part of the assessment framework, and it is well constructed in decision logic terms, with a good allocation of the burden of proof (impacts of bait use will be assumed to be occurring unless evidence to the contrary is available). Bait used in Australia’s major lobster fishery provides a big ecological and energy subsidy into the local ecosystems, and the ecological impacts have never been fully assessed. Various types of risks have been identified, but there has been little quantification. I support the concept of establishing a default table of scores for types and extent of bait use, and a linkage to related impacts of bait supply/production.

Goal: Fishery optimizes the utilization of marine resources by minimizing post-harvest loss an by efficiently using marine resources as bait. also suggest that the ‘environmental impact of bait use’ should be explicitly included in the goal statement for this Factor. Instructions: This weighting factor is addressed once for each fishery under assessment. Ascertain whether bycatch species are known or not. Both bait and dead discards are considered relative to total landings. This ratio refers to the total dead discards and/or bait use relative to total landings of all species caught in the fishery. The discard mortality rate is generally assumed to be 100% (i.e., all discards count as dead discards). Exceptions include cases where research has demonstrated high post- release survival, including invertebrates caught in pots and traps. Research that demonstrates high post- release survival for the same or similar species caught with the same or comparable gear types may qualify as showing high post-release survival. When discard mortality rates are known, multiply these rates by the amount of discards for the relevant species to determine the amount of dead discards. If the bycatch:landings ratio and/or bait use are unknown, refer to average bycatch rates for similar 38

*Underlined terms have been defined in the glossary. Ctrl + Click to view. fisheries (based on gear type, target species and/or location) as given in review papers (e.g., Kelleher 200 and Alverson et al. 1994). Bait use, if unknown, need only be addressed in cases where it is likely to be substantial relative to landings (e.g., lobster pot fishery). Err o the side of caution when there is no information.

Amount of dead discards plus bait use relative to total landings (in biomass or numbers of fish, whichever is higher)

Ratio of bait + Factor 2.4 discards/landings score

20% 1 20–40% 0.95 40–60% 0.9 60–80% 0.85 80–100% 0.8 >100% 0.75

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Criterion 2 Score and Rating

Criterion 2 Score for the stock for which you want a recommendation = Subscore Discard Rate (Factor 2.4).

• Subscore lowest score of all other assessed species caught. o Score for each species geometric mean (Factors 2.2, 2.3). Factor 2.1 is not included in the calculation; instead it modifies the score for Factor 2.2 in some circumstances (See Factor 2.2 table for more information).

Rating is based on the lowest Subscore as follows: • >3.2 Green • >2.2 and <=3.2 Yellow • <=2.2 Red

Rating is Critical if Factor 2.3 is Critical.

Criterion 3 – Management Effectiveness

Public Comment Guidance

Overview – Criterion 3 (Management Effectiveness) deals with the effectiveness of the harvest strategy, implementation, enforcement and monitoring to control fishing pressure on the managed species, as well as effectiveness of bycatch management. This criterion is composed of two factors – Harvest Strategy (3.1) and Bycatch Management Strategy (3.2), each of which is composed of numerous subfactors.

Goals for Criteria Review – While we don’t have specific proposals at this time, a goal of the criteria review is to streamline this criterion and improve its ability to distinguish better management systems. Criticisms of the current system are that it is not effectively highlighting the nuances of better management (most management systems score “moderately effective”), and is needlessly complicated, as it involves scoring numerous subfactors that rarely affect the score.

Feedback – We are interested in any and all feedback on this criterion. We have outlined specific questions for consideration under each of Factor 3.1 and Factor 3.2. In addition, please provide general comments on this criterion below. Comments: This is complex area of the decision problem. key problem is to ensure that the assessment of management effectiveness does not simply duplicate the assessment of the outcomes of management assessed in the earlier Criteria. In my view, assessing management effectiveness, in this context, needs to ensure that it is the process and structure of management that is being assessed, not the outcomes of management. This can be expressed, eg as the assessment of the likelihood that the management system as implemented will (or can) result in high quality outcomes for the preceding Criteria. This can be implemented (as partly done in SFW) by developing a list of the important process components that

*Underlined terms have been defined in the glossary. Ctrl + Click to view. should be present and operational to properly manage the fishery, and construct these into the set of parameters that boun the decision space for assessing ME. These parameters could then be assessed to determine, for example, whether the components are present in the management system, have an appropriate scope, are correctly formed, and are of sufficient quantity and quality to deliver a level of management that is consistent with the nature and scale of the fishery under assessment for the purposes of achieving high quality in the parameters to the SFW Criteria. While there is an obvious overlap, whether the outcomes are actually achieved should not be the major contributor to performance assessment grades for ME. A typical problem is that although if standards are met then this might be seen to be contributing evidence of effectiveness, but this does not protect against the situation where standards are being met even in the face of poor management, and so will not provide for the maintenance of fishery values as pressures change. One of the key attributes of a management system that should be assessed is whether the system is adequate to ensure that fishery standards can continue to be met even in the face of changing circumstances, such as in the stock or the environment externalities.

In short, probably the key roles of a management system are to ensure that a fishery attains a satisfactory standard of performance, and then to ensure that such performance is maintained, and verified, in the face of changing circumstances. In the circumstance where are fishery meets high quality sustainability standards, it may or may not be because there is a high quality system of management, and so it is important that sustainability per se is not confounded with management system quality. However, where a fishery does not meet expected standards then the management system is likely to be weak, and be in need of improvement, typically in the areas where it detects changing circumstances and needs to apply an effective response to make sure that the fishery continues to meet (or recover) high quality sustainability standards. So, the assessment of management system should be set against parameters and standards that might need to be derived from ‘best practice’, and be primarily based on the structure of the system. One way of implementing this is for the parameters used here in SFW to be re-constructed for assessment as present/absent, and the scores/grades simply assigned by statistical summary aggregation across all the component parameters (termed sub-Factors here in SFW) in a Factor. While there is always an aspect of continuous scaling, the P/A approach may be better here, assigning binary pass/fail against each sub-Factor, and to, if necessary, extend rather than contract the number of sub-Factors that would be assessed in this way.

good form of P/A is the 3 step scale presented here for implementation, which I would support (scaled as, say, fully, partly, not implemented). The Factor scores/grades could then be simply assigned by statistical summary aggregation across all the component parameters (termed sub-Factors here in SFW) in a Factor

The current SFW approach seems to closely approach this concept, based on assessing the mechanics of management system related to two areas – the harvest strategy and the bycatch strategy. However, what seems to be unclear is whether the assessment system is assessing which components of the management system are appropriate and complete, or whether the components are achieving intended outcomes. As above, this can be somewhat confusing, not least for the assessors, and there are risks of inconsistent grading outcomes. So, what this means is that in assigning a grade for ME, the assessor should be considering if the management system has components that operate in an effective way and at an adequate scale to implement the controls required for the fishery being assessed. Whether or not the management actually results in the intended outcomes, such as low levels of bycatch, would be a secondary consideration (dealt with mainly in Criterion 2). In any case the achievements of

*Underlined terms have been defined in the glossary. Ctrl + Click to view. management, in terms of outcomes for sustainability, are primarily assessed in Criteria 1 and 2, and if the achieved outcomes are also assessed in this Criterion, this could be considered as double counting specific aspects of performance. While double-counting isn’t fatal problem for an assessment structure, it does introduce the issue of balance/weighting at the Criterion level, and opens the door to some extent to ‘gaming’ of the grading system by unscrupulous operators.

Guiding principles

The fishery is managed to sustain the long-term productivity an natural ecological roles of all impacted species. Management should be appropriate (see Appendix 4 for more guidance) for the inherent vulnerability of affected marine life and should incorporate data sufficient to assess the affected species and manage fishing mortality to ensure little risk of depletion. Measures should be implemented and enforced to ensure that fishery mortality does not threaten the long-term productivity or ecological role of any species in the future.

Assessment instructions

Assess Factors 3.1 and 3.2 once for each fishery.

Factor 3.1 Harvest Strategy

Public Comment Guidance

Overview – Criterion 3.1 addresses the effectiveness of the harvest strategy for managed species, including the following subfactors: management strategy and implementation, recovery of species of concern, research and monitoring, record of following scientific advice, track record, enforcement, and stakeholder involvement.

We are seeking to streamline this criterion (ideally reducing the number of subfactors) and improve its ability to distinguish better management strategies.

Specific Proposals – We do not yet have specific proposals to streamline this category. We are proposing to eliminate the “critical” category for Factor 3.1 (see redline and strikethrough below), as it is unnecessary because any fisheries scoring a “very high concern” (1) o Factor 3.1 are automatically rated Avoid; therefore the “very high concern” is already operating as a critical factor. In addition, we propose some specific language for the “very high concern” category that would ensure that fisheries that catch species that are overfished with overfishing occurring, and do not have management in place to end overfishing and rebuild the stocks, are scored in this category. This addition is considered to be equivalent to the language previously in Factor 1.3 that scored fisheries as “Critical” in these cases.

Feedback – Please provide feedback on specific questions below. Comments: management strategy and implementation,

*Underlined terms have been defined in the glossary. Ctrl + Click to view. here I would expect to see reference to the presence of target and limit reference points that are constructed in a way and set at level expected to provide for ongoing productivity and ecosystem protection; presence of systems/tools for maintaining awareness about key parameters of the fishery (such as yield, effort, externalities, etc) and system of observers constructed to be unbiased and provide appropriate levels of sampling if the fishery is of such a scale; presence of a process for reviewing and setting quota, TAC etc to be consistent with the objectives and stakeholder concerns; and more. The construction of this in the grading table is of limited assistance to an assessor, and I would think could be expressed in a bit more detail (such as mine above) to be more explicit about what constitutes a performance threshold for the three grades of quality. recovery of species of concern, here I would expect to see thresholds for typical processes involved – is there monitoring of species of concern by independent observers, is this public domain, is there a competent authority that oversees the recovery process to ensure process integrity; …. research and monitoring, here I would have expected to see more generic statement than simply focus o stock assessment, such as information support for assessment of sensitive species, habitats, and ecological issues to verify achievement of stock and environment/ecosystem goals for the fishery record of following scientific advice, this could be deleted, or at least the performance specification needs to be modified. I fully understand and support the intent, but there are serious complexities which will make assessment likely to be inconsistent, and hence somewhat meaningless for the SFW system as a whole. First, scientific advice from stock assessment is not always the most important input to TAC etc setting, and there will be other drivers and situations where a quota/yield may be set to take account of matters not well represented, or not timely, in a stock assessment – including such things as impending climatic shifts, short term socio-economic drivers, major externalities, such as oil spills, etc. As much as I would like to think that science has a strong place in setting constraints around future harvests, practical experience shows that actual fishing constraints are often also heavily influenced by communities of stakeholders, and sometimes their perspectives are more conservative (and more appropriate) than that of typical stock assessment scenarios and bounded catch envelopes. If the science sub-factor must be kept, then it should be rephrased to something like: ‘scientific advice relevant to the fishery issues is made available for management decisions in an appropriate and timely manner’ track record, cant see any reason why this should be a sub-factor; it should be deleted. understand the point, but I would expect others in this factor to easily subsume this. enforcement, this is a key sub-factor, and probably is as important as having good reference points. Ensuring and reporting compliance with the rules, no matter how well formed they might be, and either voluntary or compulsory, is a critical capacity in fisheries of all scales and situations. Otherwise there is IUU fishing, which is clearly fatal to any concept of management. The performance scale should refer to ‘capacity to control, ensure and report compliance’ as well as to some of the key tools (VMS etc), although all of this has to be interpreted into level of appropriateness to the scale of the fishery.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. stakeholder involvement this is probably the third major component of a good fishery management system – properly engaging stakeholders into the management system. The existing gradient of performance thresholds is poorly worded in this respect.

‘included in decision-making and decisions are not made transparently’ should be included in the highest level of performance, along with consultation etc. Speaking with stakeholders, having a biased profile of stakeholders engaged, ignoring matters of concern, and not involving stakeholders in the management decisions is very weak approach to ‘involvement’, and usually leads to interest group capture with eventual outcomes that do not necessarily support productivity or ecosystem protection in the long term.

Questions:

1. Please consider the subfactors included for each factor below. Which of these seem most important? Should any be weighted more heavily in the final scoring?

2. Which, if any, may be unnecessary to score at all? Should any subfactors be combined?

3. Are there any additional considerations you would like to propose including in this criterion (this may include new subfactors, or changing the descriptions of the categories for each subfactor to include new considerations?

4. Do you have any thoughts on whether it would be more appropriate to have more than three scoring categories (currently highly effective, moderately effective or ineffective) for each subfactor, in order to provide more nuanced differentiation of management effectiveness?

5. Please provide any feedback on eliminating the “critical” concern under 3.1 (keeping in mind that all fisheries rating “very high concern” for 3.1 are already scored “Avoid”.

6. Other comments:

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Goal: Management strategy has a high chance of preventing declines in stock productivity by taking into account the level of uncertainty, other impacts o the stock, an the potential for increased pressure in the future. See Appendix 4 for more guidance.

Instructions Step 1: Assign a rating for each of the seven management subfactors using the table below: Highly effective Moderately effective Ineffective Management Fishery has highly appropriate Some effective Management strategy strategy and strategy and goals, and there management (see is insufficiently implementation is evidence (scientific or other Appendix 4 is in place, precautionary to rigorous source) that the but there is a need for protect populations strategy is being increased precaution or strategies have not implemented successfully (e.g., stronger reductions been implemented in TAC when biomass successfully For non-native species, declines, quicker reaction management policy allows for to changes in populations, For non-native prevention of further spread etc.), or management species, there are and reductions in biomass strategy is moderately known or likely over time or suppression of successful at achieving negative impacts on biomass to low levels (e.g. goals the ecosystem, below Bmsy), or includes fishery is maintained mechanisms to allow for For non-native species, in part through recovery of species impacted increases in stock size and stocking/seeding/etc., by the non-native; and further spread are and/or stock size or management is not prevented through further spread are exacerbating concern with harvesting or other not controlled by the non-native, e.g. through management measures; if harvesting or other stocking or seeding. Some any stocking or seeding strategies examples of highly effective occurs, species is already management are included in established and ongoing Appendix 4 stocking/seeding activity has been demonstrated not to contribute to growth or spread of non- native population Recovery of If overfished depleted, If overfished depleted, If overfished, species of endangered or threatened endangered or depleted endangered concern species are targeted and/or threatened species are or threatened species retained, management has a targeted and/or retained, are targeted and/or rebuilding or recovery management has a retained, and fishery

*Underlined terms have been defined in the glossary. Ctrl + Click to view. strategy in place with a high rebuilding or recovery is a substantial likelihood of success in an strategy in place whose contributor to appropriate timeframe, and eventual success is mortality of the best management practices probable or best species, management (Appendix 4 are in use to management practices to lacks an adequate minimize mortality of these minimize mortality of rebuilding or recovery species to the greatest extent “stocks of concern” are in strategy and/or possible, and harvest control use where needed and effective practices rules are in place that will are believed to be designed to limit allow for rebuilding effective mortality of these species OR OR

There are currently no Fishery is not substantial overfished depleted, contributor to mortality of endangered or threatened these species (i.e., they species targeted or retained are rated as “very low in the fishery concern” under 1.3 and/or 2.3) Scientific research The management process Some collection of data No data or very and monitoring uses an independent and up- related to stock minimal data are to-date scientific stock abundance and health is collected assessment or analysis, or collected, but data may other appropriate method be insufficient (or too that seeks knowledge related uncertain) to maintain to stock status, and this stock assessment is conducted regularly and is complete and robust, which may include both fishery-independent and appropriate fishery- dependent data and Abundance and geographic range of any non-native target species are monitored Management Management nearly always Management only Management has a record of follows scientific advice (e.g., sometimes follows track record of following does not have a track record scientific advice (e.g., only regularly exceeding scientific advice of exceeding advised TACs or sets TACs at or below the recommended TACs, otherwise disregarding recommended level half or otherwise not scientific advice) of the time) complying with scientific advice Enforcement of Regulations and agreed Enforcement and/or Enforcement and/or management voluntary arrangements are monitoring are in place to monitoring is lacking regulations regularly enforced and ensure goals are or believed to be independently verified, successfully met, although inadequate, or

*Underlined terms have been defined in the glossary. Ctrl + Click to view. including VMS, logbook effectiveness of compliance is known reports, dockside monitoring enforcement/monitoring to be poor and other similar measures may be uncertain (e.g., appropriate to the fishery regulations are enforced by fishing industry or by voluntary/honor system, but without regular independent scrutiny) Management Measures enacted by Measures enacted by Management track record management have resulted in management have not measures currently in the long-term maintenance of been in place long enough place have resulted in ecosystem integrity, including to evaluate, or the track native stock declines maintaining stock abundance record is uncertain and/or failed to allow at appropriate levels, given its depleted native role in the ecosystem and stocks to recover; or whether it is native to the have helped to ecosystem support the establishment and/or spread of a non- native species Stakeholder The management process is There is limited Stakeholders are not inclusion transparent and includes transparency, stakeholder included in decision- stakeholder input. input or consultation. making and decisions are not made transparently

Step 2 Assign a rating and a score for the fishery’s management of impacts on retained stocks (Criterion 3.1 score) based on the seven subfactors rated above. Conservation Description Score Concern: Very Low Meets or exceeds the standard of “highly effective” management for all seven 5 subfactors Low Meets or exceeds all the standard for “moderately effective “ management for 4 all seven subfactors and Meets or exceeds the standard of “highly effective” management for, at a minimum, “management strategy and implementation” and “recovery of stocks of concern” but Does not fully meet the standard for “very low concern” as defined above Moderate Meets or exceeds all the standards for “moderately effective” management for 3 all seven subfactors but Does not fully meet the standard for “low concern” as defined above High There is management in place where clear need exists, and there is not a 2 high degree of IUU fishing

*Underlined terms have been defined in the glossary. Ctrl + Click to view. and Meets or exceeds the standard for “moderately effective” management for, at minimum, “management strategy and implementation” and “recovery of stocks of concern” but Does not fully meet the standards for “moderate concern” as defined above Very High There is management in place where clear need exists, and there is not a 1 high degree of IUU fishing But No management exists when clear need for management exists in the fishery, including where catch regularly includes stocks of concern (overfished, depleted, endangered, threatened, etc.), or Substantial IUU fishing; 25% or more of the product is caught illegally. [25% is a very high quantity, and suggest should be reduced to (say) 10% or less. What often happens is that IUU fishing occurs most in places/times where fish are protected from fishing pressure or from highly effective but destructive gears, usually for a very good purpose (sensitive habitats, spawning aggregations, juvenile aggregations, etc). Admitting this form of fishing for up to 25% of the total catch seems very risky. It may be useful to define 25% as ‘critical’, or just 10% here for Very High. The other issue of course is that IUU fishing is almost always poorly quantified, often only inferred, and it may be that proxies are needed for this rather than rely on data on catch. A specialist in IUU fishing should be consulted about covert tools to estimate IUU (they exist)]. or Fishery has characteristics of “ineffective” management for “management strategy and implementation” and/or “recovery of stocks of concern”, including the case where any main species are overfished, the fishery is a substantial contributor to overfishing occurring and there is not effective management in place to end overfishing and rebuild the stock. Critical No management exists when clear need for management exists in the 0 fishery, including where catch regularly includes stocks of concern (overfished, depleted, endangered, threatened, etc.), or Substantial IUU fishing; 25% or more of the product is caught illegally.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Factor 3.2 Bycatch Management Strategy

Public Comment Guidance

We are seeking to streamline this criterion (ideally reducing the number of subfactors) and improve its ability to distinguish better management strategies.

Specific Proposal – We do not yet have specific proposals to streamline this category. We are proposing to eliminate the “critical” category for Factor 3.2 (see redline and strikethrough below), as it is unnecessary because any fisheries scoring a “very high concern” (1) on Factor 3.2 are automatically rated Avoid; therefore the “very high concern” is already operating as a critical factor.

In addition to streamlining, we are considering one major change to this factor. Currently, fisheries that are highly selective (e.g. harpoon, etc.) are scored N/A for this factor as bycatch management is unneeded. When 3.2 is scored as N/A, the entire score for management (Criterion 3) is based o the score of 3.1. This has the unintended effect of producing higher scores in some cases for less selective fisheries. We are proposing to score highly selective fisheries as “very low concern” for bycatch management, rather than N/A.

Feedback – Please provide feedback on specific questions below. Comments: haven’t had time to consider

Questions:

1. Do you agree or disagree with the proposal to score highly selective fisheries as “very low concern” for bycatch management? Why?

2. Please consider the subfactors included for each factor below. Which of these seem most important? Should any be weighted more heavily in the final scoring?

3. Which, if any, may be unnecessary to score at all? Should any subfactors be combined?

*Underlined terms have been defined in the glossary. Ctrl + Click to view. 4. Are there any additional considerations you would like to propose including in this criterion (this may include new subfactors, or changing the descriptions of the categories for each subfactor to include new considerations?

5. Do you have any thoughts on whether it would be more appropriate to have more than three scoring categories (currently highly effective, moderately effective or ineffective) for each subfactor, in order to provide more nuanced differentiation of management effectiveness?

6. Please provide any feedback on eliminating the “critical” concern under 3.2 (keeping in mind that all fisheries rating “very high concern” for 3.2 are already scored “Avoid”.

7. Other comments:

Goal: Management strategy prevents negative population impacts on bycatch species, particularly species of concern.

If the fishery has very low or no bycatch (due to highly selective gear, seasonal effort or other), score Factor 3.2 as very low concern (score of 5). as N/A for bycatch management and the management score will be based solely on Factor 3.1.

Instructions Step 1: Assign a rating for each of the four management subfactors using the table below: Highly effective Moderately effective Ineffective Management Fishery has highly effective or Strategy or Management measures strategy and precautionary strategy and implementation such as bycatch limits or implementation goals designed to understand effectiveness is bycatch reduction and minimize the impacts of under debate or techniques (Appendix 5) the fishery on bycatch species, uncertain (e.g., are in place but and there is evidence that the bycatch limits are insufficient given the strategy is being implemented imposed based on potential impacts of the successfully (e.g., there is a assumptions, but fishery (e.g., a fishery well-known track record of limits are disputed or that is heavily or fully consistently setting unsure, or bycatch exploited with no data 50

*Underlined terms have been defined in the glossary. Ctrl + Click to view. conservative bycatch limits reduction techniques collection or monitoring based o quality information are used but are of to ensure the health of and advice about bycatch, or unknown or bycatch populations), bycatch is minimized to the uncertain or greatest extent possible, effectiveness, or No management if especially for vulnerable management has not bycatch concerns are species such as sharks, been in place long possible (based on seabirds, turtles, marine enough to evaluate region and gear type) mammals and some fish its effectiveness or is but remain unknown through mitigation measures unknown) (see Appendix 5 for guidance), area closures or other bycatch reduction techniques (Appendix 5 that have been shown to be highly effective) Scientific research Adequate observer coverage or Collection of No regular collection or and monitoring effective video monitoring observer or effective analysis of data, or no coverage, data collection and video monitoring significant scientific analysis are sufficient to data exists, but research support determine that goals are being coverage or analysis met is limited Management As in Factor 3.1, but for bycatch As in Factor 3.1, but As in Factor 3.1, but for record of species (set at score for Factor for bycatch species bycatch species (set at following 3.1 if no other reason exists) (set at score for score for Factor 3.1 if no scientific advice Factor 3.1 if no other other reason exists) reason exists) Enforcement of As in Factor 3.1, but for bycatch As in Factor 3.1, but As in Factor 3.1, but for management species (set at score for Factor for bycatch species bycatch species (set at regulations 3.1 if no other reason exists) (set at score for score for Factor 3.1 if no Factor 3.1 if no other other reason exists) reason exists)

Step 2: Assign a rating and a score for the fishery’s management of impacts on bycatch species (Criterion 3.2 score) based on the four subfactors rated above. Conservation Description Score Concern Very Low Meets or exceeds the standard of “highly effective” management for all four 5 subfactors Low Meets or exceeds all the standard for “moderately effective “ management for 4 all four subfactors and Meets or exceeds the standard of “highly effective” management for, at a minimum, “management strategy and implementation” but Does not fully meet the standard for “very low concern” as defined above Moderate Meets or exceeds all the standards for “moderately effective” management for 3

*Underlined terms have been defined in the glossary. Ctrl + Click to view. all four subfactors but Does not fully meet the standard for “low concern” as defined above High There is management in place where clear need exists 2 and Meets or exceeds the standard for “moderately effective” management for, at minimum, “management strategy and implementation” but Does not fully meet the standards for “moderate concern” as defined above Very High There is management in place where clear need exists 1 but Fishery has characteristics of “ineffective” management for “management strategy and implementation” Or No bycatch management even when overfished, depleted, endangered or threatened species are known to be regular components of bycatch and are substantially impacted by the fishery Critical No bycatch management even when overfished depleted endangered or 0 threatened species are known to be regular components of bycatch and are substantially impacted by the fishery.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Criterion 3 Score and Rating

Score geometric mean (Factors 3.1, 3.2)

Rating is based on the Score as follows: • Green if >3.2 • Yellow if >2.2 and <=3.2 and neither Harvest Strategy (Factor 3.1) nor Bycatch Management Strategy (Factor 3.2) are Very High Concern or Critical • Red if <=2.2 or either the Harvest Strategy (Factor 3.1) or Bycatch Management Strategy (Factor 3.2) is Very High Concern

Rating is Critical if either or both of Harvest Strategy (Factor 3.1) and Bycatch Management Strategy (Factor 3.2) are Critical.

Criterion 4 – Impacts on the Habitat and Ecosystem

Public Comment Guidance for Criterion 4

Overview – Criterion 4 includes assessment of impacts o the seafloor habitat, and other indirect ecosystem impacts with a focus on food web/trophic impacts. Factor 4.1 scores fishery’s likely impact o the seafloor habitat based o gear type and substrate, while 4.2 allows the fishery to improve o that score due to mitigation efforts such as gear modification and spatial closures. Factor 4.3 focuses on food web impacts and the use of ecosystem-based management to avoid negative trophic impacts, as well as other ecological impacts not covered elsewhere (e.g. due to hatcheries or FADs).

Specific Proposals – 1) We are proposing a minor restructuring of Factors 4.1 and 4.2. As is, these are scored separately and each is assigned a category, but the seafloor impact is determined by the sum of the two. To emphasize that these two factors combine to create the score for impact on the habitat, we propose to rename these as Factor 4.1a and 4.1b and to assign a category (based o corresponding numerical scores) only for Factor 4.1 as a whole (the sum of 4.1a and 4.1b). Calculation of scoring would not change, therefore this would not lead to any substantive changes in ratings or reports. 2) We are proposing substantial changes to Factor 4.3, primarily to incorporate the specific guidance from the Lenfest Task Force on Forage Species. Please see the box under Factor 4.3 for more details and questions. 3) We are considering moving Factor 4.3 into new, separate criterion, as it is really separate impact from habitat effects. This would have potentially significant effects on rankings, especially due to the decision rule that one red criterion results in a Good Alternative at best and two red criteria result in an Avoid ranking. Currently, averaging factors 4.1+4.2 and 4.3 often results in a yellow for Criterion 4 overall, whereas if these two factors were assessed separately, in many cases one of them would be red. Therefore, this change would likely result in downgrades for some fisheries. comment extensively on this important matter (at points 2 and 3) in my earlier material – mainly at Criterion 1, where I try to present the argument that implementing the forage fish and EBM type arguments would best fit, addressing the direct and indirect trophic and cascading impacts issues.

Feedback – Please see questions below and provide any general feedback or comments on the proposed structural changes.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Please see Factor 4.3 for more details on the proposed changes for that factor, and please provide relevant comments in the box under 4.3. Questions:

1. Do you believe that Factor 4.3 (Ecosystem-Based Management, with a focus o Food Web) should be combined with habitat impacts in one criterion as it is currently, or should be a stand- alone criterion (or other suggestion)? As above, consider that EBM and trophic impacts should be dealt with in Criterion 1, where, if correctly constructed, the assessment would also consider the sub-point of maintaining the stock component of the target species population as part of a wider ecosystem with more natural levels of structure, function and resilience. This would be more complete and defendable interpretation of EBM, and also consider that this is an important direction for incremental improvement for wild catch fisheries, as it will enable increased security of yield in the long term, most likely (with good management) better unit returns to fishers, and provide for a recovery of the broader structure and function of ocean ecosystems.

haven’t had time to model out how this would impact existing fisheries grades, or would be influenced by the SFW system of overall grading and fail-safes. However, with any significant changes there will be some level of reputational risk incurred, and it will be important for SFW to have strategies to manage any consequences of changes to the criteria and factors that mean change (downgrading) in status of any fishery. My view is that an assessment system should respond to advances in information and technical understanding, and provide for a continuous pathway of improvement in both assessment processes and definition of the standards, even if this is perceived as an increasing standard of protection.

Comments:

Guiding principles The fishery is conducted such that impacts o the seafloor are minimized an the ecological and functional roles of seafloor habitats are maintained.

Fishing activities should not seriously reduce ecosystem services provided by any fished species or result in harmful changes such as trophic cascades, phase shifts or reduction of genetic diversity. [This aspect of the goal is not well serviced by the sub-factors – an should probably be deleted, retaining the focus on physical impacts o habitat structure. This matter should be covered to some extent by implementing Criterion 1 fully to incorporate the F issues an suitably upgraded Lenfest recommendations.

Assessment instructions

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Address Factor 4.1–4.2 for all fishing gears separately. Assess Factor 4.3 once per fishery. Factor 4.1a Physical Impact of Fishing Gear on the Habitat/Substrate Public Comment Guidance for Factors 4.1 and 4.2

Redlining below demonstrate how we propose to combine the current Factors 4.1 and 4.2 (which would be renamed 4.1a and 4.1b). This rearrangement is for communication purposes and would not materially affect the scoring for these factors in any way. Comments agree with the combination of these two matters; seems useful way to proceed.

Goal: The fishery does not adversely impact the physical structure of the seafloor or associated biological communities.

Instructions: Fishing gears that do not contact the seafloor score 5 for this criterion, and Factor 4.2 can be skipped. Use the table below to assign a score for gear impacts (Appendix 6 provides further guidance). If gear type is not listed in the table, use the score for the most similar gear type in terms of extent of bottom contact. Seafood Watch will not assess a fishery using destructive gear such as explosives or cyanide regardless of habitat type and management actions; therefore, those fishing methods are not included in the table. Where multiple habitat types are commonly encountered, and/or the habitat classification is uncertain, score conservatively according to the most sensitive plausible habitat type. Se Appendix 6 for further guidance and the methods used in developing the table below.

Conservation Description SFW Concern score None Gear does not contact bottom; fishing for a pelagic/open water species 5 Very Low Vertical line fished in contact with the bottom or fishing for a 4 benthic/demersal or reef-associated species Low Bottom gillnet, trap, bottom longline except on rocky reef/boulder and corals 3 Bottom seine (on mud/sand only) Midwater trawl that is known to contact bottom occasionally (<25% of the time) or purse seine known to commonly contact bottom Moderate Scallop dredge on mud and sand 2 Bottom gillnet, trap, bottom longline o boulder or coral reef Bottom seine (except o mud/sand) Bottom trawl (mud and sand, or shallow gravel) (includes midwater trawl known to commonly contact bottom) High Hydraulic clam dredge 1 Scallop dredge on gravel, cobble or boulder Trawl on cobble or boulder, or low energy (>60 m) gravel Very High Dredge or trawl on deep-sea corals or other biogenic habitat (such as eelgrass 0 and maerl)

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Factor 4.2 4.1b Modifying Factor: Mitigation of Gear Impacts

Goal: Damage to the seafloor is mitigated through protection of sensitive or vulnerable seafloor habitats, an limits on the spatial footprint of fishing on fishing effort.

Instructions: Assess Factor 4.2 only for fishing gear that contacts the bottom. Scores from Factor 4.2 can only improve the base score from 4.1. A high level of certainty is required to score a strong or moderate mitigation measure, e.g., good quality seafloor maps, VMS and/or observer coverage is required to document that spatial measures are effective and enforced. Further guidance can be found in Appendix 7.

Assess whether the fishery management’s efforts to mitigate the fishery’s impact on the benthic habitat fall into the categories strong, moderate, minimal or no effective mitigation, based on the table below. Factor 4.2 allows the habitat score to increase, based on the strength of mitigation measures, by the number of bonus points specified in the table.

Strong At least 50% of the representative habitat is protected from the gear type used in +1 mitigation the fishery under assessment (see Appendix 7). or Fishing intensity is sufficiently low and limited such that it can be scientifically demonstrated that at least 50% of the representative habitat is in a recovered state, based on knowledge of the resilience of the habitat and the frequency of fishing impacts from the gear type used in the fishery under assessment (see Appendix 7), or Gear is specifically designed to reduce impacts on the seafloor and there is scientific evidence that these modifications are effective in this regard and modifications are used on the majority of vessels, or Other measures are in place that have been demonstrated to be highly effective in reducing the impact of the fishing gear, which may include an effective combination of two or more “moderate” measures described below, e.g., gear modifications + spatial protection. agree with the intent, but the construction is difficult, partly because of the use of the alternatives (the ‘or’). Also, there are many definitions for ‘representative’, confounded by scale issues with spatial and taxonomic resolution of habitat parameters, and availability of measurement systems. Further, where the target species congregate and can be most efficiently fished may be the most ecologically valuable part of the fished habitat spectrum (precisely because of the congregation of the target species), so there may be variability even within a representative habitat class that should also be protected from fishing. The first two dot points seem most debatable (not the intentions, but how to operationalize in an assessment system that can be equitably and correctly applied in variety of systems/situations). haven’t had time to consider the detail, but it might be useful to provide detailed guidance to assessors to ensure they fully appreciate the complexities in these issues. agree with the use of the scoring here, effectively adding it to the total score.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. One common difficulty of using ‘or’ in a grading sub-factor is that the best level of performance will be chosen by some assessors but not others, where more than one of the options applies to a fishery (unless the system rules specify otherwise), and this opens the prospect of gaming (or at least greater contestability). Probably best resolved by being explicit about which level of performance is required to be assigned (from the spectrum of NA, best, worst) for all such sets of choices across the SFW system. In system design terms, a good way of doing this is simply to set the top level of best performance, and then reduce the performance of each assessable point of the content stepwise for lower levels of performance of the factor. Changing the points, and the form of the language, can create difficulties for assessors in making equitable interpretations on the intent within different fisheries, and at worst, can create an impression of subjectivity in an assessment. If such changes must be made, then how they relate to the top level of performance needs to be explained, and why shifts in parameter types have been needed (as guidance to assessors). Moderate Ongoing, effective measures are reducing fishing effort, intensity or spatial +0.5 mitigation footprint, or substantial proportion of all representative habitats are protected from all bottom contact and expansion of the fishery’s footprint is prohibited and vulnerable habitats are strongly protected, or Gear modifications or other measures are in use that are reasonably expected to be effective. Minimal Fishing effort or intensity is effectively controlled, but is not actively being +0.25 mitigation reduced, or Vulnerable habitats are strongly protected (e.g., in HAPCs) but other habitats are not strongly protected, or Modifications or measures anticipated to be effective are being tested or developed. No No controls on fishing intensity are in place, or +0 effective No or few efforts exist to limit the spatial extent of fishing, or mitigation No modifications anticipated to be effective are in use, or Modifications are in use but are not effective. Not Not applicable because gear used is benign and the maximum score (5) has been +0 applicable applied in 4.1a

Scoring for Factor 4.1: Impact o the Habitat

Th score for Factor 4.1 is the sum of the score for 4.1a and the score for 4.1b. Th category name for 4.1 is assigned based o score ranges, as below: Score (Sum of 4.1 and 4.2) Category <2.2 High Concern 2.2-3.2 Moderate Concern 57

*Underlined terms have been defined in the glossary. Ctrl + Click to view. >3.2 Low Concern

Factor 4.3 Ecosystem-based Fisheries Management

Public Comment Guidance for Factor 4.3

Overview – This factor assess any indirect ecological impacts not captured elsewhere in the criteria, but the focus is on potential trophic impacts, and the effectiveness of management in constraining fishing mortality to levels that are appropriate given the species’ ecological role.

Specific Proposals – At the time of writing the current criteria, there was n scientific consensus on appropriate management with regard to food web impacts. Since that time, the Lenfest Forage Fish Task Force (see http://www.oceanconservationscience.org/foragefish/ developed very detailed guidance for the management of forage fisheries. We aim to incorporate that scientific advice into our criteria. We hope, during the course of this consultation, to develop more guidance for ecosystem-based management of fisheries targeting species of exceptional importance to the ecosystem other than forage fish as well (e.g., top predators or foundational species).

The specific proposal is not yet developed into redline in the text below. The aim is to incorporate the detailed advice from the Lenfest report with the following scoring guidance:

• Fisheries on forage species (using Lenfest definition) must conform to all the recommendations of the Task Force in order to score green for this factor. • Fisheries on forage species that do not meet all the recommendations, but at least incorporate a minimum biomass threshold and a fishing mortality limit at the levels recommended by the Task Force will score yellow for this factor. • Fisheries on forage species which have biomass threshold significantly lower or fishing mortality threshold significantly higher than Lenfest recommends will score red for this factor.

Feedback – Please provide any thoughts and comments on the proposal in the box below. Questions: 1. Do you agree with the proposal to use the specific guidance in the Lenfest Forage Fish Task Force report in this factor? Yes, but I advance broader role than seems to be contemplated here. The Lenfest Task Force work has been well received, and indeed provides a strong platform for improving management systems in fisheries. The generic applicability (beyond forage species) of biomass thresholds for simple forms of management with inbuilt precaution has a number of other advantages for assessment systems, not least of which is the relatively discrete/explicit application of precaution and clear assignment of the burden of proof in relation to both direct and indirect ecosystem impacts.

argue above that the principles underlying the Lenfest case for forage species also apply more broadly to many, probably most, predator-prey interactions in fished ecosystems. To recover and maintain ocean ecosystems structure and function, I argue that the same framework should therefore be applied to management of fisheries for all species and could be both target and bycatch), using the biomass

*Underlined terms have been defined in the glossary. Ctrl + Click to view. parameters and thresholds advanced by Lenfest. There are issues about the information needed to support such an assessment approach, but there are new developments in the use of empirical proxies, and others, that could be deployed for data-poor fisheries to be assessed under the biomass thresholds approach for both productivity and ecosystem (including trophic) impacts. In any case, the limitations in knowledge-base should not be used to constrain assessments to matters that are only marginal to the main questions being posed - low power assessment that is relevant to the key issues is more appropriate in decision analysis terms and (normally) will be preferred over a high power assessment of marginally relevant set of issues.

Amongst the weaknesses that the Lenfest type of biomass thresholds approach would help to overcome is how to manage to limit direct and indirect cumulative impacts (the cascading changes in both structure and function that are imposed by heavy fishing pressure on one or more species within a region). The existing framework for managing this (the MSY frame) is almost exclusively based on single species, equilibrium models, and deterministic impacts, which apply most likely in only very few ocean systems. For example, very few management systems, or stock assessments, etc, take account (or have framework for even considering) the cumulative impacts of multiple fisheries operating in the same region.

One key aspect of generalizing the Lenfest approach would need to consider the level of the benchmarks in respect of the SFW grading scale of Black to Green. The Lenfest proposal for the benchmarks is set o the basis of productivity and limited set of ecological impacts, graded by information/knowledge quantity and quality, ie principally on a process basis. The only actual benchmarks advocated are provided as absolute limits (MaxF, Blim), and consider could be used in the SFW grading as general lowest limits to acceptable performance. For the purposes of more general application to protection of trophic and cascading and cumulative etc ecological impacts, the Lenfest benchmarks themselves are too optimistic and d not reasonably represent the extent of uncertainty in broader ecological matters across most fisheries (large and small scale, and especially those in structurally diverse ecosystems). The result of this is that the benchmarks advocated by Lenfest should be reviewed to determine if they reasonably will result in precautionary (EBM) approach to ecological protection for the ecosystems where fisheries operate. The types of uncertainty not included by Lenfest in their analysis involve the stochastic nature of almost all predator-prey relationships, the diversity of ecosystem structures, particularly species diversity, and the lack of any consideration of the ecological roles/values of different age classes of the predators or prey.

My view is that assessment benchmarks for high quality ecosystem protection should be based not only o extent of information quantity and quality (this opens the opportunity for better management, and services the aspiration without ensuring the achievement), but also o the need to cater for the level of existing uncertainty that we know exists in this area of complex cascading ecological effects in stochastic systems. As a result, in the interim (completing such review would take significant collaborative venture), for the purposes of this review of SFW and the next iteration of the assessment system, I consider that the Lenfest recommendations should be themselves used in a cautious way, until better knowledge can inform a more rigorous set of benchmarks.

This means that (for example) that the biomass thresholds should be used conservatively in the formation of the SFW grading scales. To represent current levels of uncertainty, they could be operationalised as: Lenfest low biomass limits (30% Bunfished)=Red/Yellow threshold; high biomass (80% Bunfished) = Yellow/Green threshold.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. As above, I argue that this would be best implemented not as Criterion 5, but as a part of Criterion 1. If biomass thresholds are used in both Criteria, they would be apparently (although not in reality) be used for different assessment intent purposes, and this could create significant difficulties with stakeholders. The difference of course would be based on the different proxy relationships, and associated performance grading, which in my view would be better incorporated into a single criterion involving biomass parameters established to assess performance for both production and ecological impacts purposes. This is the currently expressed intent of Criterion 1, so irrespective of Lenfest in Criterion 1 or 5, to properly achieve the expressed intent Criterion 1 needs to be significantly amended, as I discuss above.

2. If so, do you agree with the suggested guidelines for determining green, yellow, or red scores based o the Task Force recommendations? This is question about both the structure (ie information as parameter) and about performance (ie are ecosystems protected). Neither are well serviced by grading on knowledge levels alone, for the reasons above. don't consider that levels of knowledge per se will lead to a high level of precaution in fishing, or to ecologically relevant outcomes. I agree that it has important relevance, but grading on knowledge seems to be overly weighting the effectiveness of the management process in terms of potential to achieve ecological outcomes. Even complete knowledge will not necessarily result in good management/protection outcomes. So, the Lenfest guidelines need to be heavily interpreted for use in the SFW system, and I suggest above one possible way of implementation.

3. Do you have any suggestions for how to score forage fisheries for this factor? As above, I don't agree that the Lenfest approach should be constrained to forage fisheries, but even if the assessment in this area is so constrained, my comments above are also directly applicable. If the SFW cannot wholly shift to a full biomass thresholds framework, then there may be merit in (say) applying the biomass framework to fish predators, for example, as a class of target. In principle, I don't support segregating fisheries for ‘forage’ species as a special subset of fisheries, where target species deserve higher levels of protection than would species in any other fishery (unless only production objectives are being assessed, which is not the case in SFW).

4. Should Seafood Watch also address management of fisheries on top predators in this criterion (with scores based on requirements that are designed to maintain the ecological role of predators)? Is there sufficient information to effectively determine special guidelines for fisheries on predators?

argue that the approach and thresholds apply for all predator-prey interactions, if the goal of management is to maintain both productivity and the structure and function of ocean ecosystems more broadly, and if the goal of assessment is to determine if this is being achieved. Therefore my response is yes, but (as above) should be applied wider than just predators. In my view, it is feasible to apply, for the present purposes, to all predators, and could be reasonably applied to omnivores and probably grazers 60

*Underlined terms have been defined in the glossary. Ctrl + Click to view. that have large body sizes. Arguments about the lack of knowledge always have to be considered, but it is always better to have an system that has properly structured set of parameters to address the central issues to be assessed, and deal with lack of information and levels of uncertainty by applying precautionary benchmarks (which is the same framework as adopted by the Lenfest Task Force). So, even if there is not full complement of information, this should not be a reason to not adopt a properly focused assessment structure. It may result in some areas needing to be assigned with highly precautionary benchmarks, but in decision theory and for practical purposes of fishery improvement, this is a more desirable technical basis for an assessment program.

5. Do you have any suggestions for how to score fisheries for species with other ecological roles for this factor?

My discussion throughout this document is focused on all species, without limitation, although I agree that there may need to be phases of implementation, and perhaps ‘large predators’ might be a useful starting point. However, I would argue that all aspects of the SFW assessment system should be designed so that it can apply to all types of fished species, even if this needs to be phased into assessment operations.

Any Other Comments:

Goal: All stocks are maintained at levels that allow them to fulfill their ecological role and to maintain a functioning ecosystem and food web. Fishing activities should not seriously reduce ecosystem services provided by any retained species or result in harmful changes such as trophic cascades, phase shifts or reduction of genetic diversity. Even non-native species should be considered with respect to ecosystem impacts. If a fishery is managed in order to eradicate a non-native, the potential impacts of that strategy o native species in the ecosystem should be considered and rated below.

Instructions: Assign an ecosystem-based management score for the fishery, taking into consideration the ecosystem role of the main species in the fishery, and whether they are “species of exceptional importance.” Species of exceptional importance are native species that play a disproportionately important role relative to their biomass. These may include habitat-forming species, forage species, top predators, keystone species, ecosystem engineers, etc. species should not be considered of “exceptional importance” unless it clearly fits in one of these categories or there is clear evidence that it plays a keystone role in the ecosystem.

*Underlined terms have been defined in the glossary. Ctrl + Click to view. Conservation Description Score Concern Very Low There are policies in place designed to protect ecosystem functioning (e.g., a 5 substantial proportion of the fishery area is protected in no-take marine reserves which are designed and/or demonstrated to be effective for protecting ecosystem function, or there is a maximum catch or a minimum biomass based o ecosystem considerations), or an ecosystem study has been conducted and it has been scientifically proven that the fishery has no negative ecological and/or genetic impacts, and If the fishery catches “exceptional species” (see definition above), the ecological and food web impacts have been scientifically assessed and the management scheme protects enough biomass to allow these exceptional species to fulfill their ecological role (e.g., through the harvest control rule), and For fisheries with hatchery supplementation and/or use of Fish Aggregating Devices (FADs), these practices are demonstrated not to have negative ecological and/or genetic impacts, and For fisheries on non-native species, policies in place to manage the fishery and/or control the spread of the species do not have adverse effects on native species. Low Scientific assessment and management efforts to account for ecological role are 4 underway, and If the fishery catches “exceptional species”, policies are in place to protect ecosystem functioning (e.g., substantial proportion of the fishery area is protected in no-take marine reserves, or there is a maximum catch or a minimum biomass based on ecosystem considerations), though these may not be explicit or based o scientific assessment, and For fisheries with hatchery supplementation, and/or use of FADs, practices are designed to minimize or mitigate any potential negative ecological and/or genetic impacts where applicable, and For fisheries on non-native species, policies in place to manage the fishery and/or control the spread of the species do not have adverse effects on native species. Moderate The fishery does not catch “exceptional species” and scientific assessment and 3 management of ecosystem impacts are not yet underway, though they may be in planning stages or

*Underlined terms have been defined in the glossary. Ctrl + Click to view. The fishery catches “exceptional species” and lacks policies to protect the ecosystem role of these species, but scientific assessment to account for these species’ ecological roles is underway, or For fisheries with hatchery supplementation, and/or use of FADs, negative ecological and/or genetic impacts from these practices are possible and management is not designed to minimize these impacts, or For fisheries on non-native species, the policies to manage the fishery and/or control the spread of the non-native species have an unknown effect o native species. High The fishery catches “exceptional species” and there are no explicit efforts to 2 incorporate ecological role into management, or For fisheries on non-native species, policies in place to manage the fishery and/or control the spread of the species have adverse effects on native species. Very High For fisheries with hatchery supplementation, and/or use of FADs, practices are 1 in use that have been demonstrated or believed to have serious negative ecological and/or genetic impacts, or Scientifically demonstrated trophic cascades, alternate stable states, or other detrimental food web impacts are resulting from the fishery.

Criterion 4 Score and Rating

Score geometric mean (Factors 4.1+4.2, 4.3)

Rating is based on the Score as follows: • >3.2 Green • >2.2 and <=3.2 Yellow • <=2.2 Red

Rating cannot be Critical for Criterion 4. Overall Score and Final Recommendation

Public comment guidance. Overview –The final scoring system combines the individual criterion scores to produce a numerical final score from 0-5, but also applies decision rules based on the number of “high concerns”, i.e. “red” scoring criteria as outlined below.

Specifics – The following sections show how the final score and final recommendation are calculated from the individual criterion scores. It is the current philosophy of the SFW criteria that regardless of the final numerical score, if there is one red criterion (with a numerical score <=2.2), then the highest possible final recommendation is a yellow “Good Alternative”. If there 63

*Underlined terms have been defined in the glossary. Ctrl + Click to view. are two red criteria, then the overall final recommendation will be red “Avoid” regardless of the numerical score. If there is one or more “critical concerns” then the final recommendation is red “Avoid” regardless of the numerical score.

Feedback – In general the scoring system has worked, and the final recommendation is often decided by the presence of one or more red scores. The criteria are not weighted in any way at present (i.e. they all have equal weighting). This part of the criteria is critical to the appropriate final recommendation for every Seafood Watch assessment, and we welcome feedback and suggestions for improvement. Comment agree with the one Red/two Red override of the numeric scores, and I see no good reason to change it. I have reached this conclusion irrespective of whether there is or 5 Criteria (as in my discussion above) contributing to the final decision on fishery. consider this to be appropriate and clear to all stakeholders and to be reasonably consistent with the specified mission intent of SFW. understand that the application of this may downgrade the SFW status of some fisheries, but that is not a reason to not proceed to reflect the best technical understanding and available knowledge in the assessment system. Clearly, if some fisheries do have to be downgraded, then there will need to be a process implemented to manage that change, and to ensure that it is applied in such way that the fishery has an opportunity for natural justice, and, if needed, to a grace period to change fishery practices to bring it into compliance. Indeed, downgrade could be managed in such a way that a specific set of improvements could be identified for collaborative implementation to bring the fishery back into compliance within a given timeframe, and in the medium term, this may be a positive outcome for a fishery.

Final Score geometric mean of the four Scores (Criterion 1, Criterion 2, Criterion 3, Criterion 4).

The overall recommendation is as follows:

• Best Choice Final Score >3.2, and n Red Criteria, and n Critical scores

• Good Alternative Final score >2.2, and neither Harvest Strategy (Factor 3.1) nor Bycatch Management Strategy (Factor 3.2) are Very High Concern5, and n more than one Red Criterion, and n Critical scores, and does not meet the criteria for Best Choice (above)

• Avoid Final Score <=2.2, or either Harvest Strategy (Factor 3.1) or Bycatch Management Strategy (Factor 3.2) is Very High Concern5 or two or more Red Criteria, or one or more Critical scores.

5 Because effective management is an essential component of sustainable fisheries, Seafood Watch issues an Avoid recommendation for any fishery scored as a Very High Concern for either factor under Management (Criterion 3). 64

*Underlined terms have been defined in the glossary. Ctrl + Click to view. 65

*Underlined terms have been defined in the glossary. Ctrl + Click to view. <>

Glossary

[I haven’t been able to consider this is detail, and my comments are sporadic]

Adequate observer coverage: Observer coverage required for adequate monitoring depends on the rarity of the species caught, with fisheries that interact with rare species requiring higher coverage. Similarly, species that are “clumped” instead of being evenly distributed across the ocean also require higher levels of coverage. In addition, fisheries using many different gear types and fishing methods require higher levels of coverage. Bias may also be introduced if some areas, gear and seasons of a fishery are not well sampled. For these reasons, the exact level of coverage required for a particular fishery depends o the distribution of a species within the fishery as well as its associated discard and target species (Babcock and Pikitch 2003). The analyst will need to determine what level of observer coverage is adequate for the fishery of interest; coverage of 17–20% (or as high as 50% [this should be 100%] for rare species bycatch) may be required in some cases but may not be necessary in all cases. key aspect of observer coverage is how the observed effort is sampled for observation. Where observer coverage is less than 25% of the total effort, then units of observer coverage must be chosen by random sampling from the full space/time/gear spectrum of effort across the fishery. All observed data should be verifiable by independent sampling of on-board observations - remote, telephone or internet interviews with fishers does not provide a robust form of observer data. Live/storage streaming camera solutions are available and should be deployed wherever possible to keep management costs constrained.

Appropriate reference points: Determination of the appropriateness of reference points depends on two questions:

1) Is the goal appropriate? Appropriate biomass reference points are designed with the goal of maintaining stock biomass at or above the point where yield is maximized (target reference points TRPs) and safely above the point where recruitment is impaired (limit reference points LRPs). Ideally, biomass reference points aim to maintain stock biomass at levels that allow the stock to fulfill its role in the ecosystem. Fishing mortality reference points should be designed with the goal of ensuring that catch does not exceed sustainable yield and has a very low likelihood of leading to depletion of the stock in the future.

2) Is the calculation of the reference points credible? There may be concern if reference points have been lowered repeatedly or if there is scientific controversy regarding the reference points or the calculations of biomass and fishing mortality relative to reference points. Se the guidance for each type of reference point below and in Appendix 1.

Target reference point Reference points need to be evaluated on a case-by-case basis, but in general: Biomass target reference points (TRPs) should generally not be lower than BMSY or approximately B35–B40%. TRPs below about B35% require strong scientific rationale. See Appendix 1 for more details.

Limit reference point The point where recruitment would be impaired. Reference points need to be evaluated on a case-by-case basis, but in general: Biomass limit reference points (LRPs) should be no less than ½ BMSY, or ½ an appropriate target reference point such as B40%. LRPs below about B20% or BMSY require strong scientific rationale.

66 <>

Spawning potential ratio/fraction of lifetime egg production (SPR/FLEP) reference point The SPR/FLEP limit reference point should either be derived through scientific analysis to be at or above replacement %SPR for the species (the threshold level of SPR necessary for replacement) based o its productivity and S-R relationship (viz., Mace and Sissenwine 1993), or should be set at about 35–40% of LEP. A exception can be made for species with very low inherent productivity (e.g., rockfish, sharks), in which case a reference point of 50–60% of LEP is more appropriate (Mace and Sissenwine 1993, Myers et al. 1999, Clark 2002, Botsford and Parma 2005).

Bycatch: Refers to all fisheries-related injury or mortality other than in the retained catch. Examples include discards, endangered or threatened species catch, pre-catch mortality and ghost fishing. All discards, including those released alive, are considered bycatch unless there is both valid scientific evidence of high post-release survival and n documented evidence of negative impacts at the population level.

Critically endangered: A IUCN category for listing endangered species. taxon is considered “critically endangered” (CE) when it faces an extremely high risk of extinction in the wild in the immediate future, as defined by any of the relevant IUCN criteria for “critically endangered” (FAO Glossary; IUCN).

Data-moderate: Reliable estimates of MSY-related quantities are either unavailable or not useful due to life history, a weak stock-recruit relationship, high recruitment variability, etc. Reliable estimates of current stock size, life history variables and fishery parameters exist. Stock assessments include some characterization of uncertainty (Restrepo and Powers 1998; Restrepo et al. 1998).

Data-poor: Refers to fisheries for which there are n estimates of MSY, stock size, or certain life history traits. There may be minimal or no stock assessment data, and uncertainty measurements may be qualitative only (Restrepo and Powers 1998; Restrepo et al. 1998).

Data-rich: Refers to fisheries with reliable estimates of MSY-related quantities and current stock size. Stock assessments are sophisticated and account for uncertainty (Restrepo and Powers 1998; Restrepo et al. 1998).

Depleted: stock at very low level of abundance compared to historical levels, with dramatically reduced spawning biomass and reproductive capacity. Such stocks require particularly energetic rebuilding strategies. Recovery times depend o present conditions, levels of protection and environmental conditions. May refer also to marine mammals listed as “depleted” under the Marine Mammal Protection Act (FAO Glossary). Classifications of “overfished“ or “depleted“ are based on assessments by the management agency and/or FAO, but analysts can use judgment to override the classification, especially where the prior assessment may be out of date (also includes IUCN listings of “near threatened”, “special concern” and “vulnerable”). Inclusion in this classification based o designations such as “stock of concern” is determined o a case-by-case basis, as such terms are not used consistently among management agencies. Stocks should be classified as “depleted” if the stock is believed to be at a low level of abundance such that reproduction is impaired or is likely to be below an

67 <> appropriate limit reference point Marine mammals classified as “depleted” under the Marine Mammal Protection Act fall into this category, if not listed as endangered or threatened. Also includes stocks most likely (>50% chance) below the level where recruitment or productivity is impaired. Note: Official IUCN listings should be overridden by more recent and/or more specific classifications, where available (e.g., NMFS stock assessment showing stock is above target levels).

Ecological role: The natural trophic role of stock within the ecosystem under consideration in an assessment (MSC 2010).

Effective: Management or mitigation strategies are defined as “effective” if a) the management goal is sufficient to maintain the structure and function of affected ecosystems in the long-term, and b) there is scientific evidence that they are meeting these goals.

Effective mitigation or gear modification: strategy that is “effective” as defined above, either in the fishery being assessed or as demonstrated in a very similar system (See Appendix 4 and Appendix 5 for a partial list of effective mitigations; this list will be continually developed).

Endangered/threatened: Taxa in danger of extinction and whose survival is unlikely if causal factors continue operating. Included are taxa whose numbers have been drastically reduced to critical level or whose habitats have been so drastically impaired that they are deemed to be in immediate danger of extinction (FAO Fisheries Glossary). This classification includes taxa listed as “endangered” or “critically endangered” by IUCN or “threatened”, “endangered” or “critically endangered” by an international, national or state government body, as well as taxa listed under CITES Appendix I. This classification does not include species listed by the IUCN as “vulnerable” or “near threatened”.Marine mammals listed as “strategic” under the Marine Mammal Protection Act are also considered as endangered/threatened if they are listed because “based on the best available scientific information, [the stock] is declining and is likely to be listed as a threatened species under the ESA within the foreseeable future.” However, marine mammal stocks listed as “strategic” because “the level of direct human- caused mortality exceeds the potential biological removal level,” or because they are listed as “depleted” under the Marine Mammal Protection Act, are instead classified as species of concern.

Exceptional importance to the ecosystem: species that plays a key role in the ecosystem that may be disrupted by typical levels of harvesting, including: keystone species (those that have been shown or are expected to have community-level effects disproportionate to their biomass), foundation species (habitat-forming species, e.g., oyster beds), basal prey species (including krill and small pelagic forage species such as anchovies and sardines), and top predators, where the removal of a small number of the species could have serious ecosystem effects. Species that do not fall into any of these categories but that have been demonstrated to have an important ecological role impeded by harvest (e.g., studies demonstrating trophic cascades or ecosystem phase shifts due to harvesting) shall also be considered species of exceptional importance to the ecosystem (Paine 1995; Foley et al. 2010).

FishBase vulnerability: Cheung et al. (2005) used fuzzy logic theory to develop an index of the intrinsic vulnerability of marine fishes. Using certain life history parameters as input variables—maximum length, age at first maturity, longevity, von Bertalanffy growth parameter K, natural mortality rate, fecundity, strength of spatial

68 <> behavior and geographic range—together with heuristic rules defined for the fuzzy logic functions, fish species were assigned to either very high, high, moderate or low level of intrinsic vulnerability. In this framework, intrinsic vulnerability is also expressed with a numerical value from 1 to 10 with 10 being most vulnerable. This index of intrinsic vulnerability was then applied to over 1300 marine fish species to assess intrinsic vulnerability in the global fish catch (Cheung et al. 2007). FishBase, the online global database of fish, includes this numerical “vulnerability score” o the profiles of fish species for which this analysis has been conducted (Froese and Pauly 2010).

Fluctuating biomass: • If stock is trending upwards (based on the most recent assessment) and has just recently exceeded the target reference point (TRP), it can be rated as Very Low Concern. If stock is not trending but is truly fluctuating around the TRP (exceeding in some years and falling short in others, but with no clear trend), it can be rated as Very Low Concern. However, if stock is fluctuating around the limit reference point (LRP), it cannot be considered Very Low Concern. • If stock is trending downward and is currently below the TRP, the rating can be no better than Low Concern. • If stock is below the LRP, it is considered High Concern.

Fluctuating fishing mortality:

• If F is trending downwards, or was previously above FMSY (or suitable proxy) but has recently gone below FMSY (in the most recent assessment), fishing mortality should be rated as Low Concern.

• If F is fluctuating around FMSY or if F has been consistently below FMSY and has just recently (in the latest assessment) risen above FMSY for just this one year (potentially due to management error or new stock assessment and the consequent adjustment in reference points or estimates), fishing mortality should be rated as Moderate Concern.

• If F is trending upwards and has just risen above FMSY fishing mortality should be rated as High Concern unless there is a substantial plan to bring F back down. Such plan would need to differ substantially from the existing harvest control rules (HCR), as those evidently did not keep at a sufficiently low level.

Highly appropriate management strategy: Management that is appropriate for the stock and harvest control rules takes into account major features of the species’ biology and the nature of the fishery. Such a management strategy incorporates the precautionary approach while also taking uncertainty into account and evaluating stock status relative to reference points, as these measures have been shown to be robust (modified from MSC 2010). As an example, if management is based on Total Allowable Catch, these limits are set below MSY and/or scientifically advised levels, accounting for uncertainty, and lowered if B

Likelihood: Highly likely: 70% chance or greater, when quantitative data are available; may also be determined by analogy from similar systems when supported by limited data from the fishery and n scientific controversy exists (modified from MSC 2010 and based on guidance from MSC FAM Principle 2). Examples of high likelihood for fishing mortality:

69 <>

• All estimates of fishing mortality are within 70% or greater (e.g., 80% or 95%) confidence intervals, or all estimates of F from all scientifically feasible models and assumptions, are below FMSY or equivalent.

• Estimate of is at 75% of sustainable levels of F (such as FMSY or less, or estimates of catch are < 75% MSY if the fishery is at or above BMSY (if is fluctuating, see “fluctuating F” for more guidance). • Exploitation is very low compared to natural mortality, mortality from other sources or population size

Likely: 60% chance or greater, when quantitative data are available; may also be determined according to expert judgment and/or plausible argument (modified from MSC 2010 and based on guidance from MSC FAM Principle 2).

Probable: Greater than 50% chance; can be based o quantitative assessment, plausible evidence or expert judgment. Examples of “probable” occurrence for fishing mortality:

There may be some uncertainty or disagreement among various models; fishing mortality may be above 75% of sustainable level and/or catch may be above 75% of sustainable catch level (e.g., MSY) for stocks at BMSY.

Historic high: Refers to near-virgin biomass, unfished biomass, or highest recorded biomass, if biomass estimates predate the start of intensive fishing. If fishery has been historically depleted and then rebuilt, the rebuilt biomass is not considered “historic high” even though it may be higher than historical levels.

Inherent vulnerability: stock’s vulnerability to overfishing based on inherent life history attributes that affect the stock’s productivity and may impede its ability to recover from fishing impacts. Marine finfish are considered highly vulnerable when their FishBase vulnerability score falls into the categories high “vulnerability”, “high to very high vulnerability”, or “very high vulnerability” (scores of 55 or above) based on the FishBase vulnerability score (derived from the formula in Cheung et al. 2005). All sea turtles, marine mammals, and seabirds are considered “highly vulnerable”. Marine invertebrates’ vulnerability is based o the average of several attributes of inherent productivity.

One of the first key papers on this subject (Musick 1999) summarizes the results of an American Fisheries Society (AFS) workshop on the topic and offers proposed low, medium and high “productivity index parameters” (for marine fish species) based on available life history information: the intrinsic rate of increase r, the von Bertalanffy growth function k, fecundity, age at maturity and maximum age. Notably, although a species’ intrinsic rate of increase is identified as the most useful indicator, it is also difficult to estimate reliably and is often unavailable (Cheung et al. 2005). To enable more timely and less data-intensive and costly identification of vulnerable fish species, Cheung et al. (2005) used fuzzy logic theory to develop an index of the intrinsic vulnerability of marine fishes based on life history parameters: maximum length, age at first maturity, longevity, von Bertalanffy growth parameter K, natural mortality rate, fecundity, strength of spatial behavior and geographic range (input variables). The index also uses heuristic rules defined for the fuzzy logic functions to assign fish species to one the following groups: very high, high, moderate or low level of intrinsic vulnerability.

70 <>

In this framework, intrinsic vulnerability is also expressed by a numerical value between 1 and 10 with 10 being most vulnerable. This index of intrinsic vulnerability was then applied to over 1300 marine fish species to assess intrinsic vulnerability in the global fish catch (Cheung et al. 2007). FishBase, the online global database of fish, uses the numerical value from this index as a “vulnerability score” o the profiles of fish species for which it has been evaluated (Froese and Pauly 2010).

Large portion of the stock is protected: At least 50% of the spawning stock is protected, for example through size/sex/season regulations or the inclusion of greater than 50% of the species’ habitat in marine reserves. Future guidance will improve the integration of marine reserve science into the criteria, based on ongoing research.

Main species: species is included in the assessment as a main species if: • The catch of the species in the fishery under assessment composes >5% of that fishery’s catch, or • The species is >1% of that fishery’s catch and the fishery causes >5% of the species’ total mortality across all fisheries, or • The species is <1% of that fishery’s catch and the fishery causes >20% of species’ total mortality across all fisheries, or • The species is overfished, depleted, stock of concern, endangered, threatened, IUCN Near Threatened, US MMPA strategic species, and/or subject to overfishing and the fishery causes >1% of species’ total mortality across all fisheries. • If there are no other “main species” (based on the above guidance) besides the one assessed under criterion 1, but the total catch of other discarded and retained species is >5% (i.e. catch of criterion species is <95% of total), assess the top 3 species by volume of catch (if there are only 1-2 other species caught, assess those species).

Managed appropriately: Management uses best available science to implement policies that minimize the risk of overfishing or damaging the ecosystem, taking into account species vulnerability along with scientific and management uncertainty, and applies appropriate levels of precaution at all relevant points in the management system to maintain stock productivity and provide for minimum ecological impacts on ecosystems that may be affected by fishing.

Negative genetic or ecological effects: Hatchery supplementation of fisheries may have negative ecological effects, including competition between hatchery-raised and wild fish, as well as genetic effects, including reduced fitness of populations through genetic introgression and reduced genetic diversity (Kostow 2009; Araki and Schmid 2010). Other fishing practices may have additional detrimental genetic or ecological effects; for example, drifting Fish Aggregating Devices (FADs) may act as “ecological traps” that mislead fish into making maladaptive habitat choices that reduce their fitness (Hallier and Gaertner 2008). Unfortunately, these impacts are generally poorly understood for most fisheries, but where impacts are likely (i.e., where FADs are in use or where large-scale hatchery supplementation exists, as in most salmon fisheries), ecosystem-based management should include measures specifically designed, and in the best cases, proven to reduce these impacts.

71 <>

Negligible: Mortality is insignificant or inconsequential relative to a sustainable level of total fishing mortality (e.g., MSY or PBR); less than or equal to 5% of a sustainable level of fishing mortality.

No management: fishery with no rules or standards for regulating fishing catch, effort or methods. Management does not need to be enforced through government regulation or official management agencies but may also include voluntary action taken by the fishery, as long as there is general compliance.

Non-native species:

non-native organism is an organism occurring outside its natural past or present range and dispersal potential, including any parts of the organism that might survive and subsequently reproduce, whose dispersal to the non-native area is caused by direct human action (e.g., introduced through ballast water or intentional translocation of organisms to a new area, but not including indirect anthropogenic effects, such as range shifts due to climate change). Modified from Falk-Petersen et al 2006.

Overfished: stock is considered “overfished” when exploited past an explicit limit where abundance is considered too low to ensure safe reproduction. In many fisheries, the term “overfished” is used when biomass has been estimated below a biological reference point used to signify an “overfished condition”. The stock may remain overfished (i.e., with a biomass well below the agreed limit) for some time even though fishing mortality may have been reduced or suppressed (FAO Glossary). Classification as “overfished“ or “depleted“ (including IUCN listing as “near threatened”, “special concern” and “vulnerable”) is based o evaluation by the management agency and/or FAO, but an analyst can use judgment to override this classification, especially where the classification may be out of date as long as there is scientific justification for doing so. Inclusion in the “overfished” category based o designations such as “stock of concern” are determined on case-by-case basis, as such terms are not used consistently among management agencies. Stocks should be classified as “overfished” if the stock is believed to be at such low level of abundance that reproduction is impaired or is likely to be below an appropriate limit reference point Marine mammals classified as “depleted” under the Marine Mammal Protection Act also fall into this category if not listed as endangered or threatened. Stocks that are most likely (>50% chance) below the level where recruitment or productivity is impaired are also considered “overfished”. Note: Official IUCN listings should be overridden by more recent and/or more specific classifications where available (e.g., NMFS stock assessment showing that a stock is above target levels).

Overfishing: generic term used to refer to level of fishing effort or fishing mortality such that reduction of effort would, in the medium term, lead to an increase in the total catch; or, rate or level of fishing mortality that jeopardizes the capacity of a fishery to produce the maximum sustainable yield on a continuing basis. For long-lived species, overfishing (i.e., using excessive effort) starts well before the stock becomes overfished. Overfishing, as used in the Seafood Watch® criteria, can encompass biological or recruitment overfishing (but not necessarily economic or growth overfishing). • Biological overfishing Catching such a high proportion of one or all age classes in a fishery as to reduce yields and drive stock biomass and spawning potential below safe levels. In a surplus production model, biological overfishing occurs when fishing levels are higher than those required for extracting the

Maximum Sustainable Yield (MSY) of a resource and recruitment starts to decrease.

72 <>

• Recruitment overfishing When the rate of fishing is (or has been) high enough to significant reduce the annual recruitment to the exploitable stock. This situation is characterized by a greatly reduced spawning stock, decreasing proportion of older fish in the catch and generally very low recruitment year after year. If prolonged, recruitment overfishing can lead to stock collapse, particularly under unfavorable environmental conditions. • Growth overfishing: Occurs when too many small fish are being harvested too early through excessive fishing effort and poor selectivity (e.g., excessively small mesh sizes), and the fish are not given enough time to grow to the size at which maximum yield-per-recruit would be obtained from the stock. Reduction of fishing mortality among juveniles, or their outright protection, would lead to an increase in

yield from the fishery. Growth overfishing occurs when the fishing mortality rate is above Fmax (in a yield- per-recruit model). This means that individual fish are caught before they have a chance to reach their maximum growth potential. Growth overfishing, by itself, does not affect the ability of a fish population to replace itself. • Economic overfishing Occurs when a fishery is generating no economic rent, primarily because an excessive level of fishing effort is applied in the fishery. This condition does not always imply biological overfishing.

(FAO Glossary; NOAA 1997)

Precautionary approach: The precautionary approach involves the application of prudent foresight. Taking account of the uncertainties in fisheries systems and considering the need to take action with incomplete knowledge, the precautionary approach requires, inter alia: (i) consideration of the needs of future generations and avoidance of changes that are not potentially reversible; (ii) prior identification of undesirable outcomes and measures to avoid or correct them promptly; (iii) initiation of any necessary corrective measures without delay and on a timescale appropriate for the species’ biology; (iv) conservation of the productive capacity of the resource where the likely impact of resource use is uncertain; (v) maintenance of harvesting and processing capacities commensurate with estimated sustainable levels of the resource and containment of these capacities when resource productivity is highly uncertain; (vi) adherence to authorized management and periodic review practices for all fishing activities; (viii) establishment of legal and institutional frameworks for fishery management within which plans are implemented to address the above points for each fishery, and (ix) appropriate placement of the burden of proof by adhering to the requirements above (modified from FAO 1996).

Productivity is maintained/not impaired: Fishing activity does not impact the stock, either through reduced abundance, changes in size, sex or age distribution, or reduction of reproductive capacity at age, to a degree that would diminish the growth and/or reproduction of the population over the long-term (multiple generations).

Productivity–susceptibility analysis (PSA): Productivity-susceptibility analysis was originally developed to assess the sustainability of bycatch levels in Australia’s Northern Prawn fishery (Patrick et al. 2009), and has since been widely applied to assess vulnerability to fishing mortality for as number of fisheries worldwide. Productivity-susceptibility analysis is used by NOAA and the Australian Commonwealth Scientific Industrial Research Organization (CSIRO) to inform fisheries management. It also constitutes the basis of the risk-based framework used to evaluate data-poor fisheries under the Marine Stewardship Council Fishery Assessment Methodology

73 <>

(MSC FAM). The PSA approach allows the risk of overfishing to be assessed for any species based on predetermined attributes, even in the most data-poor situations.

The exact sets of productivity and susceptibility attributes vary between PSA methodologies, and different weighting of attributes can be employed based o relative contextual importance. Additionally, scoring thresholds can vary depending o the context in which PSA is employed. In the US methodology, productivity is defined as the capacity for a stock to recover once depleted, which is largely a function of the life history characteristics of the species. Generally, productivity attributes are similar to the life history parameters used for the above index of intrinsic vulnerability.

While PSA analysis is a widely accepted approach for evaluating risk of overexploitation of fished species, for the purposes of Seafood Watch assessments it is useful to separate the productivity attributes—which are intrinsic to a species and neither dependent on nor influenced by fishery practices—from the susceptibility attributes. Fisheries may influence the susceptibility of impacted stocks through the choice of gear, bait species, hook design, mesh size, area or seasonal closures, and other management measures. In addition, where detailed information o fishing mortality (e.g., estimates of or harvest rates) is available, these data provide more complete picture of the fishery impact that the susceptibility attributes are designed to predict. Therefore, under the revised Seafood Watch® criteria, the “inherent vulnerability” score will be derived from the FishBase vulnerability score, which address only those characteristics intrinsic to the species and equivalent to the productivity attributes considered under PSA. Susceptibility attributes will be separately considered as part of the evaluation of fishing mortality when more specific data are not available.

Because the FishBase vulnerability index has not been evaluated for application to marine invertebrates (Cheung, pers comm., 2011), we propose the use of “productivity” attributes from the PSA methodology used by the MSC, with adjustments to account for particular aspects of marine invertebrate life history (see Criterion Factors 1.1 and 2.1 for guidance).

Changes to PSA resilience attributes made to increase applicability to invertebrates include:

• Removal of “average maximum size” and “average size at maturity” attributes, as body size has been demonstrated not to correlate with extinction risk for marine invertebrates (McKinney 1997, Finnegan et al. 2009). • Incorporation of fecundity in the average only if fecundity is low, as strong evidence suggests that high fecundity does not necessarily correlate with low vulnerability; however, low fecundity does seem to correspond with high vulnerability (Dulvy et al. 2003, Cheung et al. 2005). • Removal of the “trophic level” attribute. Anecdotally, trophic level does not appear to be a strong predictor of extinction risk for marine invertebrates. Some of the most vulnerable marine invertebrates are at the lowest trophic level (e.g., abalone). Additionally, in review of Canadian fishes, trophic level was not found to be a strong predictor of extinction risk in marine fishes (O'Malley 2010; Pinsky et al. 2011). • Incorporation of density dependence, particularly the existence of depensatory dynamics, or Allee effects, for small populations. Allee effects may have a profound effect on the resilience of marine invertebrates to fishing mortality (Hobday et al. 2001, Caddy 2004).

Reasonable timeframe (for rebuilding):

74 <>

Dependent on the species’ biology and degree of depletion, but generally within 10 years, except in cases where the stock could not rebuild within 10 years even in the absence of fishing. In such cases, reasonable timeframe is within the number of years it would take the stock to rebuild without fishing, plus one generation, as described in Restrepo et al. (1998).

Recruitment is impaired: Fishing activity impacts the stock—either through reduced abundance, changes in size, sex or ag distribution, or reduction of reproductive capacity at age—to a degree that will diminish the growth and/or reproduction of the population over the long-term (multiple generations); or, the stock is below an appropriate limit reference point if one is defined.

Regularly monitored: Fishery-independent surveys of stocks, or other reliable assessments of abundance, are conducted at least every three years.

Reliable data: Data produced or verified by an independent third party. Reliable data may include government reports, peer- reviewed science, audit reports, etc. Data are not considered reliable if significant scientific controversy exists over the data, or if data are old or otherwise unlikely to represent current conditions (e.g., survey data is several years old and fishing mortality has increased since the last survey).

Species of concern: Species about which management has some concerns regarding status and threats, but for which insufficient information is available to indicate a need to list the species as endangered. In the U.S., this may include species for which NMFS has determined, following a biological status review, that listing under the ESA is "not warranted," pursuant to ESA section 4(b)(3)(B)(i), but for which significant concerns or uncertainties remain regarding their status and/or threats. Species can qualify as both "species of concern" and "candidate species" (http://www.nmfs.noaa.gov/pr/glossary.htm#s). In addition, marine mammal stocks listed as “strategic” under the Marine Mammal Protection Act are classified as species of concern. The terms “species of concern” or “stock of concern” are used similarly by other federal and state management bodies.

Stakeholder input: stakeholder is an individual, group or organization that has an interest in, or could be affected by, the management of a fishery (modified from MSC 2010).Stakeholder input may include: involvement in all key aspects of fisheries management from stock assessment and setting research priorities to enforcement and decision-making. In addition, stakeholders may take ownership of decisions and greater responsibility for the wellbeing of individual fisheries (Smith et al 1999). Effective stakeholder engagement requires that the management system has consultation processes open to interested and affected parties and that roles and responsibilities of the stakeholders are clear and understood by all relevant parties (modified from MSC 2010).

Stock: self-sustaining population that is not strongly linked to other populations through interbreeding, immigration or emigration. single fishery may capture multiple stocks of one or multiple species. Stocks can be targeted or non-targeted, retained or discarded, or some combination thereof (e.g., juveniles are discarded and adults are retained).

75 <>

Ideally, the management unit of “stock” should correspond to the biological unit. However, often the fisheries management unit of “stock” may not be the same as the biological unit. For example the biological population may be much more widespread than the ‘stock’ being fished, and impacts on the ‘stock’ need to be interpreted in the context of ecological structure and function beyond the spatial boundaries of the ‘stock’. This commonly occurs when a ‘stock’ is an aggregation of the target species, but which are normally found widely spread and in abundances too low to be targeted for fishing. This context is always important for ecological purposes, since pressures are commonly great in areas of low density of a naturally distributed species, and this may provide a driver for adaptation capacity of the population. If multiple biological stocks are managed as one, and there is insufficient information to assess the stock status of each biological stock, the management unit is assessed. This situation detracts from the fishery’s “quality of information” score, as it makes it impossible to assess individual stocks’ health. If management occurs on a finer scale than biological stocks so that multiple management unit stocks compose one interbreeding population, the health and abundance of the biological stock should be assessed as whole based on information aggregated across the management units. The effectiveness of management can be assessed at the finest scale for which meaningful and verifiable differences in management practice exist.

Substantial contributor: fishery is substantial contributor to impacts affecting population, ecosystem or habitat if the fishery is a main contributor, or one of multiple contributors of a similar magnitude, to cumulative fishing mortality. Examples of fishery that is not substantial contributor include: (a) catch of the species is a rare or minor component of the catch in this fishery and the fishery is small contributor to cumulative mortality, relative to other fisheries. However, if there has been a jeopardy determination for that stock in the fishery being assessed it should be considered a substantial contributor regardless of this definition.

Substantial proportion of habitat: Refers to a condition when at least 20% of each representative habitat (where representative habitats can be delineated by substrate, bathymetry, and/or community assemblages), within both the range of the targeted stock(s) and the regulatory boundaries of the fishery under consideration (i.e., within the national EEZ for the fishery under consideration), is completely protected from fishing with gear types that impact the habitat in that fishery.

Susceptibility (low/moderate/high) stock’s capacity to be impacted by the fishery under consideration, depending o factors such as the stock’s likelihood to be captured by the fishing gear. The susceptibility score is based on tables from MSC’s Productivity- Susceptibility Analysis framework (see Appendix 2). Examples of low susceptibility include: low overlap between the geographic or depth range of species and the location of the fishery; the species’ preferred habitat is not targeted by fishery; the species is smaller than the net mesh size as an adult, is not attracted to the bait used, or is otherwise not selected by fishing gear; or strong spatial protection or other measures in place specifically to avoid catch of the species.

Sustainable level (of fishing mortality): level of fishing mortality that will not reduce stock below the point where recruitment is impaired, i.e., above F reference points, where defined. The limit reference points should be around either FMSY or F35–40% for moderately productive stocks; low productivity stocks like rockfish and sharks require F in the range of F50–60% or lower. Higher values require a strong scientific rationale. The reference points are limit reference points, so buffers should be used to ensure that fishing mortality does not exceed these levels. Where F is unknown but

76 <>

MSY is estimated, fishing mortality at least 25% below MSY is considered sustainable level (for fisheries that are at or above BMSY).

Up-to-date data/stock assessment: Complete stock assessments are not required every 1–3 years, but stocks should be regularly monitored at least every 1–3 years, and stock assessments should be based on abundance and fishing mortality data not more than three years old. Data may be collected by industry, but analysis should be independent.

Very limited area: Fishing (with damaging gear, when assessing Criterion D) is limited to no more than 50% of each representative habitat (where representative habitats can be delineated by substrate, bathymetry, and/or community assemblages) within both the range of the targeted stock(s) and the regulatory boundaries of the fishery under consideration (e.g., the national EEZ for the fishery under consideration).

Very low levels of exploitation (e.g., experimental fishery): Fishery is under-exploited or is being conducted experimentally to collect data or gauge viability, such that exploitation rates are far below sustainable yields (e.g., 20% or less of sustainable take). Alternatively, when n other information is available, exploitation levels may be considered very low if a fishery falls into the “low” category for all “susceptibility” questions under Productivity-Susceptibility Analysis (see Appendix 2 for more details on scoring susceptibility).

77 <>

References

Alverson, D.L. Freeberg, M.H., Pope, J.G., Murawski, S.A. 1994. A global assessment of fisheries bycatch and discards. FAO Fisheries Technical Paper. No. 339. Rome, FA. 233p

Araki, H, and C. Schmid. 2010. Is hatchery stocking help or harm? Evidence, limitations and future directions in ecological and genetic surveys. Aquaculture 308: S2-S11

Auster, P.J. 2001. Defining thresholds for precautionary habitat management actions in a fisheries context. North American Journal of Fisheries Management 21:1-9.

Australian Department of Agriculture, Fisheries and Forestry 2007. Commonwealth Fisheries Harvest Strategy: Policy and Guidelines. December 2007. Available at: http://www.daff.gov.au/_media/documents/fisheries/domestic/HSP-and-Guidelines.pdf

Babcock, E., E. Pikitch, and C. Hudson. 2003. “How much observer coverage is enough to adequately estimate bycatch?” Oceana, Washington D.C.

Botsford, L. W., and A. M. Parma. 2005. Uncertainty in Marine Management. Pages 375-392 in E. A. Norse and L. B. Crowder, editors. Marine Conservation Biology: The Science of Maintaining the Sea's Biodiversity. Island Press, Washington, DC.

Caddy, J. F. 2004. Current usage of fisheries indicators and reference points, and their potential application to management of fisheries for marine invertebrates. Canadian Journal of Fisheries and Aquatic Science 61:1307-1324.

Chaffee C., S. Daume S, L. Botsford, D. Armstrong, and S. Hanna. 2010. Oregon Dungeness Crab Fishery Final Report with Certification Decision. Ver. 4, 2 October 2010. Available at: http://www.msc.org/track-a-fishery/certified/pacific/oregon-dungeness-crab/assessment-downloads- 1/ODC_V4_Final-Report_27Oct2010.pdf

Cheung, W. W. L., T. J. Pitcher, and D. Pauly. 2005. fuzzy logic expert system to estimate intrinsic extinction vulnerabilities of marine fishes to fishing. Biological Conservation 124:97-111.

Cheung, W. W. L., R. Watson, T. Morato, T. J. Pitcher, and D. Pauly. 2007. Intrinsic vulnerability in the global fish catch. Mar. Ecol. Prog. Ser. 333:1-12.

Chuenpagdee, R., L. E. Morgan, et al. (2003). "Shifting gears: Assessing collateral impacts of fishing methods in US waters." Frontiers in Ecology and Environment 1(10): 517-524.

Clark, W.G. 1991. Groundfish exploitation rates based o life history parameters. Canadian Journal of Fisheries and Aquatic Sciences 48: 734-750.

Clark W. G. 2002. F35% revisited ten years later North American Journal of Fisheries Management 22:251-257.

78 <>

DFO. 2009. fishery decision-making framework incorporating the Precautionary Approach. Fisheries and Oceans Canada. Available at: http://www.dfo-mpo.gc.ca/fm-gp/peches-fisheries/fish-ren-peche/sff- cpd/precaution-eng.htm

Dulvy, K., Y. Sadovy, and J. D. Reynolds. 2003. Extinction vulnerability in marine populations. Fish and Fisheries 4:25-64.

FAO. 1996. Precautionary Approach to Capture Fisheries and Species Introductions. FAO Technical Guidelines for Responsible Fisheries, 2: 54 p.

FAO Fisheries Glossary. Accessed December 8, 2010. Available at: http://www.fao.org/fi/glossary/

Falk-Petersen, J., T. Bøhn & O. T. Sandlund. 2006. On the numerous concepts in invasion biology. Biological Invasions 8:1409–1424.

Field, J., Cope, J. & Key, M. 2010. descriptive example of applying vulnerability evaluation criteria to California nearshore finfish species. Managing Data-Poor Fisheries: Case Studies, Models & Solutions, 1, 235–246.

Finnegan, S., S. C. Wang, A. G. Boyer, M. E. Clapham, Z. V. Finkel, M. A. Kosnik, M. Kowalewski, R. A. J. Krause, S. K. Lyons, C. R. McClain, D. McShea, P. M. Novack-Gottshall, R. Lockwood, J. Payne, F. Smith, P. A. Spaeth, and J. A. Stempien. 2009. No general relationship between body size and extinction risk in the fossil record of marine invertebrates and phytoplankton. in GSA Annual Meeting, Portland, OR.

Foley, M.M., Halpern, B.S., Micheli, F. et al. 2010. Guiding ecological principles for marine spatial planning. Marine Policy 34: 955-966.

Froese, R., T.A. Branch, A. Proelß, M. Quaas, K. Sainsbury, and C. Zimmermann. 2010. Generic harvest control rules for European fisheries. Fish and Fisheries: doi:10.1111/j.1467-2979.2010.00387.x

Froese, R., and D. Pauly. 2010. FishBase. World Wide Web electronic publication. www.fishbase.org.

Fuller, S.D., C Picco, et al. 2008. How we fish matters: Addressing the ecological impacts of Canadian fishing gear. Ecology Action Centre, Living Oceans Society, and Marine Conservation Biology Institute. 25pp.

Goodman, D., M. Mangel, G. Parkes, T. Quinn, V. Restrepo, T. Smith and K. Stokes. 2002. Scientific Review of The Harvest Strategy Currently Used in The Bsai and Goa Groundfish Fishery Management Plans, North Pacific Fishery Management Council, Anchorage, AK. 153 p. Available at: http://www.fakr.noaa.gov/npfmc/misc_pub/f40review1102.pdf

Hall, M. 1998. An ecological view of the tuna-dolphin problem: Impacts and tradeoffs. Reviews of fish biology and fisheries (8):1-34

Hallier, J.P. and D. Gaertner. 2008. Drifting fish aggregation devices could act as an ecological trap for tropical tuna species. Marine Ecology Progress Series 353:255-264.

79 <>

Hobday, A., M. J. Tegner, and P. L. Haaker. 2001. Over-exploitation of broadcast spawning marine invertebrate: Decline of the white abalone. Reviews in Fish Biology and Fisheries 10:493-514.

Johannes, R. E. 1998. The case for data-less marine resource management: Examples from tropical nearshore fisheries. Trends in Ecology and Evolution 13: 243–246.

Jones, J.B. 1992. Environmental impact of trawling on the seabed: a review. New Zealand Journal of Marine and Freshwater Research. 26: 59-67

Kaiser, M.J., J. S.Collie, S. J Hall, S. Jennings, I. R. Poiner. 2001. Impacts of fishing gear on marine benthic habitats. Reykjavik Conference o Responsible Fisheries in the Marine Ecosystem. Reykjavik, Iceland. ftp://ftp.fao.org/fi/document/reykjavik/pdf/12kaiser.PDF Kelleher, K. 2005. Discards in the world’s marine fisheries. An update. FAO Fisheries Technical Paper. No. 470. Rome, FAO. 131p.

Kostow, K. 2009. Factors that contribute to the ecological risks of salmon and steelhead hatchery programs and some mitigating strategies. Reviews in Fish Biology and Fisheries 19:1, 9-31.

Lindeboom, H. J., and Groot, S. J. de (eds) 1998. The effects of different types of fisheries on the North Se and Irish Se benthic ecosystems. NIOZ-Rapport 1998-1, RIVO-DLO Report C003/98: 404 pp.

Lindholm, J. B., P. J. Auster, M. Ruth, and L. S. Kaufman. 2001. Modeling the effects of fishing, and implications for the design of marine protected areas: juvenile fish responses to variations in seafloor habitat. Conservation Biology 15:424–437.

Mace, P.M. and M.P. Sissenwine. 1993. How much spawning per recruit is enough? pp 101–118 in S.J. Smith, J.J. Hunt and D.Revered (eds.) Risk Evaluation and Biological Reference Points for Fisheries Management. Canadian Special Publication of Fisheries and Aquatic Sciences 120. National Research Council of Canada.

Marine Stewardship Council (MSC). 2010. Fisheries Assessment Methodology and Guidance to Certification Bodies, version 2.1, including Default Assessment Tree and Risk-Based Framework. Marine Stewardship Council, London, UK. 120 p. Available at: http://www.msc.org/documents/scheme- documents/methodologies/Fisheries_Assessment_Methodology.pdf

McKinney, M. L. 1997. Extinction vulnerability and selectivity: Combining ecological and paleontological views. Annual Review of Ecology, Evolution and Systematics 28:495-516. MSC. 2009. Marine Stewardship Council Fisheries Assessment Methodology and Guidance to Certification Bodies Including Default Assessment Tree and Risk-Based Framework

Myers R. A., Bowen K. G., and Barrowman N. J. 1999. Maximum reproductive rate of fish at low population sizes Canadian Journal of Fisheries and Aquatic Sciences 56:2404-2419.

Musick, J. A. 1999. Criteria to define extinction risk in marine fishes. Fisheries 24:6-14.

New Zealand Ministry of Fisheries. 2008. Harvest Strategy Standard for New Zealand Fisheries. 25 p. Available at: http://fs.fish.govt.nz/Doc/16543/harveststrategyfinal.pdf.ashx

80 <>

NOAA. 1997. NOAA Fisheries Strategic Plan. 48p. Available at: http://www.nmfs.noaa.gov/om2/nmfsplan.pdf

O'Malley, S. L. 2010. Predicting vulnerability of fishes. Masters Thesis. University of Toronto, Toronto.

Paine, R.T. 1995. "A Conversation on Refining the Concept of Keystone Species". Conservation Biology 9 (4): 962–964.

Patrick, W. S., P. Spencer, O. Ormseth, J. Cope, J. Field, D. Kobayashi, T. Gedamke, E. Cortés, K. Bigelow, W. Overholtz, J. Link, and P. Lawson. 2009. Use of productivity and susceptibility indices to determine stock vulnerability, with example applications to six U.S. fisheries. U.S. Department of Commerce, NOAA Tech. Memo. NMFS-F/SPO-101, Seattle, WA.

Pinsky, M.L., Jensen, O.P., Ricard, D., Paulumbi, S.R. 2011. Unexpected pattersn of fisheries collapse in the world’s oceans. PNAS: 108 (20) 8317-8322

Restrepo, V.R. and J. E. Powers. 1998. Precautionary control rules in US fisheries management: specification and performance. ICES Journal of Marine Science 56: 846–852.

Restrepo, V.R., G.G. Thompson, P.M. Mace, W.L. Gabriel, L.L. Low, A.D. MacCall, R.D. Methot, J.E. Powers, B.L. Taylor, P.R. Wade and J.F. Witzig. 1998. Technical Guidance on the Use of Precautionary Approaches to Implementing National Standard 1 of the Magnuson-Stevens Fishery Conservation and Management Act. NOAA Technical Memorandum NMFS-F/SPO-137. 54pps.

Roughgarden, J. and F. Smith. 1996. Why fisheries collapse and what to do about it. Proc. Nat. Acad. Sci., (USA), 93:5078-5083

81 Version October 8, 2014

Appendices

Appendix – Further guidance on interpreting the health of stocks and fishing mortality

The tremendous variability among fisheries makes it impossible to define specific appropriate reference points that would be applicable to all assessed fisheries for the purposes of maintaining production in the long term Instead, criteria are based on the commonly accepted management goal that target biomass should be at or above the point where yield is maximized, and management should ensure a high probability that biomass is at or above a limit reference point (where recruitment or productivity of the stock would be impaired). Three common types of reference points are MSY-based, SPR-based, and ICES reference points. However, other reference points may be used in some fisheries, and should be evaluated in accordance with the management goal articulated above. [this is a good summary of the usual situation, but it is important to recognize the context – these reference points and default positions are constructed (correctly) to respond to concerns about the capacity of fish stocks for continued production. The theoretical constructs and practical management experience reflected in this ‘guidance’ does not relate to the ecological functions of the fish stock or population in ecosystems, and explicitly avoids a discussion of how reference points for production maximization (or optimization) relate to reference points for maintaining the ecological role of a fish stock within a natural ecosystem. Key issues include the use of deterministic constructs when we know that ecosystems are stochastic, the lack of predator-prey linkages, dependencies and cascading effects in developing reference points to protect the role of a fish stock in ecosystems, and considerably more. While these production reference points, if constructed and applied with appropriate levels of precaution that reflect the various uncertainties, will likely protect a fish species from becoming extinct, they make only a small contribution to protecting and supporting natural structure and function of ecosystems. Extinction is very low performance bar, and it is likely that the natural role of species that would be otherwise found in high abundance in an ecosystem (high enough to support target fishery) will be severely (ecologically) depleted by management to achieve maximum/optimum production benchmarks, even though it may not become extinct. Further, In regions of the world where there is high structural diversity (such as the tropics and many subtropics), single species production benchmarks take almost n account of the cumulative impact of fishing for multiple species in local area or region, and so the cumulative impact of multiple fisheries, each independently managed to achieve maximum/optimum yield and maintained at levels of about 30 to 40% of unfished biomass with truncated age structures, is likely to be large and very significant for ecosystem structure and function in those areas. Here, I suggest that this discussion be clearly articulated as a discussion about production and yield, and the caveats about ecosystem structure and function noted, so that assessors are correctly guided in their overall assessments.]

MSY-based reference points While the concept of MSY is far from perfect, MSY-based biomass and fishing mortality reference points are commonly used in some of the most well managed fisheries around the world.When applied appropriately, these reference points are an important tool for maintaining stock productivity in the long term. However, without properly accounting for scientific and management uncertainty, maintaining a stock at BMSY (the biomass corresponding to MSY) and harvesting at MSY runs a high risk of unknowingly either overshooting MSY or allowing biomass to drop below BMSY without reducing harvest rates and thus inadvertently overharvesting (Roughgarden and Smith 1996; Froese et al. 2010). The risk

*Underlined terms have been defined in the glossary. Ctrl Click to view. of impacts from inadvertent overharvesting increases with increased uncertainty and with increased inherent vulnerability of the targeted stock. To account for these interactions, the guidance provided for assessing stock health and fishing mortality is based on MSY reference points but requires high scientific confidence that biomass is above target levels and that fishing mortality is below MSY.

Proxies for BMSY are acceptable if shown to be conservative relative to BMSY for that stock, or if they fit within the guidelines for appropriate target level*. Where BMSY or proxy reference points are not known or are not applicable, the stock/population health criteria can be interpreted using relevant indicators that are appropriate as targets and safe limits for abundance of the species (e.g., escapement relative to escapement goals can be evaluated in lieu of biomass relative to limit reference points).

SPR-based reference points In the absence of stock assessments and MSY-based reference points, the stock health can be evaluated based on CPUE, trends in abundance and size structure, and/or simple, easy to calculate reference points such as Fraction of Lifetime Egg Production (FLEP) (equivalent to Spawning Potential Ratio, SPR). Other data-poor or alternative assessment techniques that provide evidence that stocks are healthy (i.e., productivity and reproduction are not impaired) may be used in place of reference points. Guidelines may be added on a continuing basis, but they must be demonstrated to be accurate indicators of stock health. Guidance for assessing stock health and fishing mortality for data-poor stocks is based on O’Farrell and Botsford (2005) and Honey et al. (2010) but can be addressed on a case-by- case basis with a goal of ensuring that data-poor fisheries are held to the same standard of likelihood as data-rich fisheries when stocks are above a level where recruitment would be impaired and fishing mortality is at or below a sustainable level of harvest.

• Examples of evidence that a stock is above the point where recruitment or productivity is impaired. i.e. an appropriate limit reference point, include: o the current lifetime egg production (LEP) or spawning per recruit (SPR) is above an appropriate SPR or Fraction of Lifetime Egg Production (FLEP)-related reference point; o spawning potential is well protected (e.g., females are not subject to mortality, and it can be shown or inferred that fertilization is not reduced); o quantitative analyses conducted by fishery scientists under transparent guidelines indicate sufficient stock • Strong, quantitative scientific evidence from the fishery under consideration is required to consider stock “very low concern”. When limited data are available from the fishery, analogy with similar systems, qualitative expert judgments and/or plausible arguments may be used to consider the stock “low concern”. • Use of CPUE requires the absence of hyperstability, that CPUE is proportional to abundance (or adjusted), and that there have been no major changes in technology. • If relying on CPUE rather than a LEP-based reference point, trends in size structure are also needed to ensure the fishery is not reducing stock productivity by depleting the relative proportion of large individuals. • The LEP can be estimated from length frequency data from both unfished (or marine reserve) and current populations, and does not require catch-at-age data. Reference points based on FLEP should be considered limit reference points.

83 For “very low concern” ratings, there must be no evidence that productivity has been reduced through fisheries-induced changes in size or age structure, size or age at maturity, sex distribution, etc. SPR- based and MSY-based reference points should account for these changes as they are based on productivity of the stock rather than simple abundance. If the metric considers abundance only, or if there is evidence that productivity has been reduced through shifts in age, size or sex distributions, the stock cannot be rated higher than “low concern”.

ICES reference points The ICES reference points Fpa, Flim, Bpa, and Blim are not equivalent to MSY-based reference points.In fact, comparisons have demonstrated that Fpa is typically above FMSY and Bpa is typically below BMSY, such that MSY-based reference points are generally more conservative (ICES 2010). In many cases, Bpa is well below BMSY and even below 1/2BMSY (Kell et al. 2005). Therefore, guidance for evaluating stock health using Bpa and fishing mortality using Fpa is conservative, accounting for the difference between these reference points and MSY-based reference points. ICES plans to transition to an MSY-based approach by 201 (ICES 2010). If B>Bpa or F

Proxies For many fisheries, FMSY and BMSY are unknown, and proxies are often used. Most commonly, biomass proxies are based o the percent of unfished or virgin biomass (B0).Fishing mortality proxies are often based o spawning potential ratio (SPR).

Commonly used and acceptable biomass reference points are typically 35–40% of B0 for most stocks (Clark 1991; NZ Ministry of Fisheries 2008). This target may vary according to stock productivity; however, justifications for lower target levels are often based o assumptions about “steepness6“ that may be highly uncertain or poorly understood. It is now recognized that stock targets lower than approximately 30-40% of B are increasingly difficult to justify (NZ Ministry of Fisheries 2008). For these targets to be considered appropriate reference points, solid scientific justification is required. In addition, stocks reduced to this target level or below (equivalent to removing more than 60–70% of the stock’s biomass) would be unlikely to achieve the ecosystem-based management goal of allowing a stock to fulfill its ecological role and should be scored accordingly under ecosystem-based management.

Alternatively, when unfished biomass cannot be estimated, proxy biomass reference points may be based o the equilibrium biomass achieved using appropriate fishing mortality reference points, as described below.

A large body of scientific literature addresses appropriate fishing mortality reference points based on spawner biomass per recruit (SPR). Ideally, these should be shown through scientific analysis to be at or above replacement %SPR (the threshold level of SPR necessary for replacement) for the species, based on its productivity and S-R relationship (viz. Mace and Sissenwine 1993). However, for many species this analysis will not be available. In these cases, guidance is based o the conclusions of numerous analyses demonstrating that, in general, F35-40% (the fishing mortality rate that reduces the SPR to 35–40% of unfished levels) is appropriate for species with moderate vulnerability, while a more conservative fishing mortality rate of about F50-60% is needed

6 Steepness is a key parameter of the Beverton-Holt spawner-recruit model that is defined as the proportion of unfished recruitment produced by 20% of the unfished spawning biomass. Steepness is difficult to estimate, and the calculation of reference points is often very sensitive to estimates of steepness.

84 for highly vulnerable species such as rockfish and sharks (Botsford and Parma 2005; Mace and Sissenwine 1993; Clark 2002; Myers et al. 1999; Goodman et al. 2002).

Evaluating Fishing Mortality

Evaluation of fishing mortality should reflect the mortality caused by the fishery, but in the context of whether cumulative impacts on the species (including mortality from other fisheries) are sustainable. When determining whether a fishery is a substantial contributor err on the side of caution. Unknown or missing data are grounds for classification as a substantial contributor.

Reference points Generally, species should be managed with reference points that fit the definition of a sustainable level of fishing mortality and/or an appropriate SPR or Fraction of Lifetime Egg Production (FLEP)-related reference point Species that are not commercially fished or managed but make up non-target catch in the fishery will generally not have reference points defined. In lieu of reference points, these stocks should be evaluated relative to a level of mortality scientifically shown not to lead to depletion of the stock. For species with high vulnerability, the reference point must be demonstrated to be appropriate for that species’ biology. As a rule of thumb, F40% is not precautionary enough for high vulnerability species; F50% or lower is more appropriate when using SPR-based proxies.

ICES reference points Because analysis has shown that the ICES reference point Fpa is typically above FMSY ICES stocks using Fpa as reference point must be rated more conservatively than stocks using FMSY If F > Fpa, rate the stock as “high concern”. If F < Fpa, rate the stock as “moderate concern,” unless there is additional evidence that is below sustainable level such as Fmsy.

Data-poor stocks When no reference points are available (i.e., in data-poor fisheries), fishing mortality can be evaluated based o the likelihood that management actions and characteristics of the fishery constrain fishing mortality to acceptable levels. For example, fishing mortality could be considered low concern if the fishery has a low likelihood of interacting with a non-target species due to low overlap between the species range and the fishery, or due to low gear selectivity for the species (resulting in low susceptibility; see below). Fishing mortality o target or non-target species may be considered a low concern if there is a very low level of exploitation as in an experimental emerging fishery, or if large portion of the stock’s key habitat is protected in no-take reserves that have been shown to effectively protect the species. We hope to incorporate the results of ongoing research to provide more specific guidance o the amount of habitat that should be protected in marine reserves. In the interim, however, in order to score fishing mortality as very low concern in the absence of reference points or other information about fishing mortality rates, at least 50% of all representative habitats should be protected in Marine Protected Areas where the assessed stock experiences n fishery mortality.

Appendix 2 – Susceptibility attribute from th MSC FAM

Where no information on fishing mortality is available in a data-poor fishery, susceptibility may be used as proxy (see Factor 1.3). The concept of susceptibility is based o the “Susceptibility” attributes from the Productivity-Susceptibility Analysis (PSA) used in the MSC Certification Requirements document (2012). The susceptibility attributes and scoring guide from the MSC are given below. “Selectivity”

85 should be scored only once for the appropriate gear type; where n cut-offs for a particular gear type are provided, the analyst shall develop similar selectivity tables that are appropriate for the gear and shall include a justification for the factors used and cut-offs selected in these cases.

Instructions for use: for each species in each fishery, score each attribute of susceptibility below. The final susceptibility score is the arithmetic mean of each attribute, and is considered “low” susceptibility overall if the final score is <=1.5, moderate if between 1.5 and 2.5, and high if >=2.5

Low susceptibility (low Medium susceptibility High susceptibility (high risk, risk, score: 1) (medium risk, score: 2) score: 3) Areal overlap – Overlap of the fishing effort with a species distribution of the stock. <10% overlap 10–30% overlap >30% overlap Vertical overlap – The position of the stock/species within the water column relative to the fishing Low overlap with fishing Medium overlap with gear. gear fishing gear High overlap with fishing gear Selectivity for set gillnets Selectivity is the potential of gear to capture or Length at maturity < Length at maturity is 1-2 Length at maturity >2 times retain the mesh size, or >5 m in times mesh size or 4-5 m mesh size, to say, 4 m in species length in length length Selectivity for hooks a. Does not eat bait Defined by typical (e.g. diet specialist), filter weights of the feeder (e.g. basking species caught relative to shark), small mouth (e.g. the breaking strain of the sea horse). a. Large species, with a. Bait used in the fishery is snood, the gaffing Most robust scoring adults rarely caught, but selected for this type of method used in the attribute. juveniles captured by species, and is a known diet fishery, and by diet of hooks. preference (e.g. squid bait potential species b. Species with used for swordfish), or capacity to break b. Species with important in wild diet. Scores for hook line when hooked capacity to break susceptibility may be (e.g. large toothed snood when being b. Species unable assigned using the whales, and sharks). landed. to break snood categories to the right. If when being landed there are conflicting c. Selectivity known c. Selectivity known answers, e.g. Low on to be low from to be medium from c. Selectivity known point 1 but medium on selectivity selectivity to be high from selectivity point 2, the higher risk analysis/experiment analysis/experiment analysis/experiment score shall be used. (e.g. <33% of fish (e.g. 33-66% of fish (e.g. >66% of fish encountering gear encountering gear are encountering gear are selected) selected). are selected) Selectivity for Traps/Pots a. Cannot physically a. Can enter and easily a. Can enter, but cannot Scores for trap enter the trap (e.g. too escape from the trap, but easily escape from the susceptibility may be big for openings, sessile is attracted to the trap trap, and is attracted to assigned using the species, wrong shape, (e.g. does eat the bait, or either the bait, or the categories to the right. If etc.). trap is attractive as habitat provided by the trap.

86 there are conflicting habitat) answers, e.g. b. Can enter and easily b. Species regularly found in Low on point 1 but escape from the trap, b. Can enter, but cannot the trap medium on point 2, the and no incentive to enter easily escape from the higher risk score shall be the trap (does not eat trap, and no incentive to used. bait, trap is not enter the trap (does not attractive as habitat, eat bait, trap is not etc.) attractive as habitat, etc.)

c. Species occasionally found in the trap. Post-capture mortality (scores vary by fishery)

The chance that, if captured, a species would be released in condition that would permit Evidence of post-capture Retained species, or majority subsequent survival release and survival Released alive dead when released

87 Appendix 3 – Matrix of bycatch impacts by gea type

The matrix shown in this appendix is used to determine the relative impact of a fishery on bycatch species of various taxa for fisheries where species and amounts of bycatch are not available or are incomplete. The matrix represents typical relative impacts of different fishing gear on various taxa based o the best available science. The values in the matrix were developed initially by averaging the findings of two studies that ranked the relative ecological impacts of fishing gear (Fuller et al. 2008; Chuenpagdee et al. 2003). Some values in the matrix have been updated based o a survey of scientific experts on bycatch from around the world to increase the global relevance of the matrix.

The findings of the studies used to construct this matrix were pulled from literature searches, fisheries data and expert opinion. In general, these studies ranked the severity of fishing gear impacts as shown in this table (in order of severity):

Chuenpagdee et al 2003 Fuller et al 2008 Bottom trawl Bottom trawl Bottom gillnet Bottom gillnet Dredge Dredge Midwater gillnet Bottom longline Pot and traps Midwater trawl Pelagic longline Pot and trap Bottom longline Pelagic longline Midwater trawl Midwater gillnet Purse seine Purse seine Hook and line Hook and line Dive Harpoon

Because these studies were based o fisheries operating in Canadian and United States waters, we also conducted a review of literature and expert opinion o bycatch severity by gear type from different regions of the world.Some of the initial values from Fuller et al. (2008) and Chuenpagdee et al. (2003) were adjusted accordingly. These changes are intended to better reflect the array of bycatch issues that occur using the same gear types in different regions of the world. For example, we increased the potential severity of bycatch of seabirds for trawlers operating in regions where there is a high occurrence of albatross and petrel interaction based o studies that found (1) seabirds may collide with net monitoring (net sonde) cables and trawl warps, and (2) seabirds can become entangled in the net (while attempting to feed) when the trawl is at the surface during setting and hauling (Weimerskirch et al. 2000; Wienecke and Robertson 2002; Sullivan 2006; Bull 2007; Abraham et al. 2008).

In addition, we increased the potential severity of midwater trawl impacts o marine mammals based o the study of DuFresne et al. (2007), which found fishing depth (midwater) to be the most important predictor of common dolphin bycatch. Another study by Fertl and Leatherwood (1997) found more cetaceans caught in midwater than bottom trawls. Also, pinniped species have been shown to use trawl fisheries for food and may become trapped in nets and drown (Fertl & Leatherwood 1997; Morizur et al. 1999; Rowe 2007). For example, New Zealand fur seals are attracted to the physical presence of fishing vessels and are thought to become caught when the net is being shot or during hauling (Baird & Bradford 2000).

88 Bycatch severity for biogenic habitats (coral and sponges) by gear type was determined by averaging the values given in Fuller et al. (2008) and Chuenpagdee et al. (2003). In Chuenpagdee et al. (2003), this category was named “biological habitat” and in Fuller et al. (2008) it was called “coral and sponges.” We did not change these values because it is likely that gear types that contact the bottom have the same potential for severe impacts throughout the world’s oceans. Impacts from fishing on the benthos occur o virtually all continental shelves worldwide (Watling 2005).

We increased the number of trawl types from only bottom and midwater (used in both Fuller et al. (2008) and Chuenpagdee et al. (2003)) to also include bottom trawl categories for tropical/subtropical fish, tropical/subtropical shrimp, coldwater fish, and coldwater shrimp. Shrimp trawls are not designed to drag along the bottom and herd fish, so they receive a lower impact score in the matrix for finfish bycatch. The turtle bycatch score was increased for tropical trawls to account for the severe threats to turtles posed by bottom trawling in tropical waters worldwide.

Other changes to the findings of Fuller et al. (2008) and Chuenpagdee et al. (2003) include separating the different purse seine techniques into FAD/log sets, dolphin/whale sets and unassociated school sets based o the variable bycatch rates found in study by Hall (1998). Hall (1998) found that log (FAD) sets have the overall greatest bycatch for some species, followed by school sets and dolphin sets. Turtle bycatch potential was derived from research by Wallace and Lewison (2010).

Bottom seines or demersal seines (including Danish seines, Scottish fly-dragging seines and pair seines) were not included in the Fuller et al. (2008) and Chuenpagdee et al. (2003) studies because these gear types are not commonly used in the U.S. or Canada. Like purse seines, these gear types are used to encircle school of fish, but they are operated in contact with the seafloor. A study by Palsson (2003) compared haddock discards among three demersal gear types in Icelandic waters and found fish bycatch to be lowest in Danish seines when compared with demersal trawls and longline gear. Danish seines targeting benthic fish species can incidentally catch non-target species such as flatfish, cod, and haddock (Icelandic Ministry of Fisheries 2010). Alverson et al. (1996) found that Danish seines generally fell into a low-moderate bycatch group of gear, with lower bycatch ratios than the majority of gear types, including bottom trawls, longlines and pots, but with higher bycatch than pelagic trawls and purse seines. Based on these findings, the bycatch score of Danish seines was estimated from the score for purse seines with an increase in the effects on shellfish to account for Danish seines being operated on the seafloor, an increase in the effect on finfish to account for greater bycatch of benthic fish such as flatfish, cod and haddock, and decrease in the effect on forage fish, which are typically pelagic.

89 Unknown bycatch matrix

Highest impacts receive a score of 1 and lowest impacts receive a score of 5. Key: B = Bottom, P = Pelagic, M = Mid-water, BTF = Bottom tropical fish, BTS = Bottom tropical shrimp, BCF = Bottom coldwater fish, BCS = Bottom coldwater shrimp, PF = Purse FAD/log (tuna), PD = Purse dolphin/whale (tuna), PU = Purse unassociated (tuna), Pot = Pot and trap, HD = Harpoon/diver, TP = Troll/pole and line

Longline Gillnet Trawl Dredge Seine Other B P B M B BTF BTS BCF BCS M B P PF PD PU Pot HD TP Benthic Inverts 4.5 5 3 5 2 2 2 2 2 5 1 3 5 5 5 5 3.5 5 5 Finfish 2 2 2 1 2 1 2 2.5 2 1 3 2 4 2 3 2 3.5 5 3 Forage Fish 5 4 2 2 3 2 2 2 2 1 5 3 3 2 3 2 4 5 4 Sharks 3 2 2 2 2 2 2 2 2 2 5 3.5 3.5 2 5 3 5 5 3.5 Mammals 4 4 1 1 2 1 4 4 4 2 5 3.5 3.5 5 3 5 4 5 5 Seabirds (within albatross range) 1 1 1 1 2 2 2 2 2 1 5 4 4 4 4 4 4 5 4.5 Seabirds (outside albatross range) 5 3 3 1 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Turtles (within turtle range) 4 1 2 1 4 1 1 5 5 4 3 4 4 3 5 4 5 5 4 Corals and other biogenic habitats 3 5 2 5 1 1 1 1 1 5 1 1 5 5 5 5 3.5 5 4.5

90 Additional guidance for unknown bycatch species

Sea turtles – all endangered/threatened: Se Wallace et al. (2010) for global patterns of marine turtle bycatch. In addition, a global program, Mapping the World's Sea Turtles, created by the SWOT (State of the World's Sea Turtles) database is a comprehensive global database of sea turtle nesting sites around the world. The SWOT map is highly detailed and can be customized, allowing location filters and highlights of both species and colony size with variously colored and shaped icons. This map together with the paper by Wallace et al. (2010) can help to determine if the fishery being assessed has potential interactions with sea turtles.

Sharks, marine mammals and seabirds: Identify whether the fishery overlaps with any endangered/threatened or overfished species and err on the side of caution if species-specific and geographic information is inconclusive. For example, if shark populations are data deficient, err on the side of caution and rate as “overfished” or “depleted”.

Sharks Select “overfished” or “depleted”, when data deficient or select “endangered/threatened” when data exist to support this (see Camhi et al. 2009). Globally, three-quarters (16 of 21) of oceanic pelagic sharks and rays have an elevated risk of extinction due to overfishing (Dulvy et al. 2008). See Camhi et al. (2009) for geographic areas, IUCN status and conservation concerns by shark species. Table illustrates additional resolutions, recommendations and conservation and management measures by RFMO for sharks.Additional region and species-specific shark conservation information associated follows Table in list format (Camhi 2009; Bradford 2010).

Marine mammals: The global distribution marine mammals and their important conservation areas are given by Pompa et al. (2011), who also used geographic ranges to identify 2 key global conservation sites for all marine mammal species (123) and created range maps for them (Figure 1; Table 2; Pompa et al. 201 and supplt.). [I agree that this should be included – important information- but it is also important to observe that species richness per se is only one attribute of the issue of marine mammal conservation/protection, and that even total protection of sites of high richness is not likely to be adequate to ensure ongoing protection/recovery of marine mammal populations. It is not sufficient to prioritise fishery interactions in places of high mammal species richness, although I agree that the mapped distribution does alert assessors to a higher likelihood of interactions even if it does not greatly contribute to quantifying the risks of fishery interactions in a specific fishery with a specific mammal population.]

Seabirds Figures 2 and 3 illustrate the distribution of threatened seabirds throughout the world (Birdlife International 2011). Also see Birdlife International (2010) to locate Marine Important Bird Areas (MIBA). Albatross are the most highly threatened family, with all 22 species either globally threatened or near threatened. The penguins and shearwaters/gadfly petrels also contain high proportion of threatened species (Birdlife International 2010). [ditto, for the same reasons]

Teleost fish species Unknown (a Moderate Concern) unless data exist that would indicate an alternative rating.

91 Table 1. Active resolutions, recommendations, and conservation and management measures by RFMO for sharks. Table from Camhi et al.(2009). a ICCAT = International Commission for the Conservation of Atlantic Tunas; NAFO = North Atlantic Fisheries Organization; GFCM = General Fisheries Commission for the Mediterranean; SEAFO = South East Atlantic Fisheries Organization; IATTC = Inter-American Tropical Tuna Commission; WCPFC = Western and Central Pacific Fisheries Commission; IOTC = Indian Ocean Tuna Commission; CCAMLR = Commission for the Conservation of Antarctic Marine Living Resources. The weight of recommendations and resolutions varies by RFMO. For example, all ICCAT recommendations are binding, whereas resolutions are not.

Ocean/ Res/Rec Title Main actions RFMOa/Y No.b ear Atlantic, ICAAT 1995 Res. 95-2 Resolution by ICCAT on cooperation with the • Urges members to collect species-specific data on biology, bycatch and trade in shark Food and Agriculture Organization of the species and provide these data to FAO United Nations with regard to study on the status of stocks and by-catches of shark species 2003 Res. 03-10 Resolution by ICCAT on the shark fishery • Requests all members to submit data on shark catch, effort by gear, landings and trade in shark products • Urges members to fully implement a NPOA 2004 Rec. 04-10 Recommendation by ICCAT concerning the • Requires members to annually report shark catch and effort data • Requires full utilization conservation of sharks caught in association • Bans finning • Encourages live release • Commits to reassess shortfin mako and blue with fisheries managed by ICCAT sharks by 2007 • Promotes research on gear selectivity and identification of nursery areas 2005 Rec. 05-05 Recommendation by ICCAT to amend • Requires annual reporting of progress made toward implementation of Rec. 04-10 by Recommendation 04-10 concerning the members • Urges member action to reduce North Atlantic shortfin mako mortality conservation of sharks caught in association with fisheries managed by ICCAT 2006 Rec. 06-10 Supplementary recommendation by ICCAT • Acknowledges little progress in quantity and quality of shark catch statistics • Reiterates concerning the conservation of sharks caught call for current and historical shark data in preparation for blue and shortfin mako in association with fisheries managed by ICCAT assessments in 2008 2007 Rec. 07-06 Supplemental recommendation by ICCAT • Reiterates mandatory data reporting for sharks • Urges measures to reduce mortality of concerning sharks targeted porbeagle and shortfin mako • Encourages research into nursery areas and possible time and area closures • Plans to conduct porbeagle assessment no later than 2009 2008 Rec. 08-07 Recommendation by ICCAT on the • Urges live release of bigeye thresher sharks to the extent practicable • Requires bigeye conservation of bigeye thresher sharks shark catches and live releases be reported (Alopias superciliosus) caught in association with fisheries managed by ICCAT Atlantic, NAFO 2009 Mgt. Conservation and management of sharks • Requires reporting of all current and historical shark catches • Promotes full utilization • Measure Bans finning • Encourages live release • Promotes research on gear selectivity and

92 Article 17 identification of nursery areas Atlantic, SEAFO 2006 Conservatio Conservation measure 04/06 on the • Same provisions as ICCAT Rec. 04-10, except does not include stock assessments n measure conservation of sharks caught in association 04/06 with fisheries managed by SEAFO Med., GFCM 2005 GFCM/2005/ Recommendation by ICCAT concerning the • Same provisions as ICCAT Rec. 04-10 3 conservation of sharks caught in association with fisheries managed by ICCAT 2006 GFCM/2006/ Recommendation by ICCAT to amend • Same provisions as ICCAT Rec. 05-05 8(B) Recommendation [04-10] concerning the conservation of sharks caught in association with fisheries managed by ICCAT Indian, IOTC 2005 Res. 05/05 Concerning the conservation of sharks caught • Requires members to report shark catches annually, including historical data • Plans to in association with fisheries managed by IOTC provide preliminary advice on stock status by 2006 • Requires full utilization and live release • Bans finning • Promotes research on gear selectivity and to ID nursery areas 2008 Res. 08/01 Mandatory statistical requirements for IOTC • Requires members to submit timely catch and effort data for all species, including members and cooperating non-contracting commonly caught shark species and less common sharks, where possible parties (CPCs) 2008 Res. 08/04 Concerning the recording of catch by longline • Mandates logbook reporting of catch by species per set, including for blue, porbeagle, fishing vessels in the IOTC area mako and other sharks Pacific, IATTC 2005 Res. C-05-03 Resolution on the conservation of sharks • Promotes NPOA development among members • Work with WCPFC to conduct shark caught in association with fisheries in the population assessments • Promotes full utilization • Bans finning • Encourages live release Eastern Pacific Ocean and gear-selectivity research • Requires species-specific reporting for sharks, including historical data 2006 Res. C-04-05 Consolidated resolution on bycatch • Requires prompt release of sharks, rays and other non-target species • Promotes research (REV 2) into methods to avoid bycatch (time-area analyses), survival rates of released bycatch and techniques to facilitate live release • Urges members to “provide the required bycatch information as soon as possible” Pacific, WCFPC 2008 Cons. & Mgt. Conservation and management measure for • Urges members to implement the IPOA and report back on progress • Requires annual Measure sharks in the Western and Central Pacific reporting of catches and effort • Encourages live release and full utilization • Bans finning 2008-06 Ocean for vessels of all sizes • Plans to provide preliminary advice on stock status of key sharks by (replaces 2010 2006-05)

93 Southern, CCAMLR 2006 32-18 Conservation of sharks • Prohibits directed fishing of sharks • Live release of bycatch sharks

94 Additional shark information and citations (Bradford 2010)

• In the Gulf of Mexico, Baum and Myers (2004) found that between the 1950s and the late- 1990s, oceanic whitetip and silky sharks (formerly the most commonly caught shark species in the Gulf of Mexico) declined by over 99 and 90%, respectively. • In the Northwest Atlantic, Baum et al. (2003) estimated that scalloped hammerhead, white, and thresher sharks had declined by over 75% between the mid 1980s and late 1990s. The study also found that all recorded shark species in the Northwest Atlantic, with the exception of mako sharks, declined by over 50% during the same time period. • Myers et al. (2007) reported declines of 87% for sandbar sharks, 93% for blacktip sharks, 97% for tiger sharks, 98% for scalloped hammerheads, and 99% or more for bull, dusky, and smooth hammerhead sharks along the Eastern seaboard since surveys began along the coast of North Carolina in 1972. • The International Union for the Conservation of Nature (IUCN) has declared that “32% of all pelagic sharks and rays are Threatened.” The IUCN has declared another 6% to be Endangered, and 26% to be Vulnerable. • In the Mediterranean Sea, Ferretti et al. (2008) found that hammerhead, blue, mackerel, and thresher sharks have declined between 96 and 99.99% relative to their former abundance levels. • Ward and Myers (2005) report a 21% decline in abundance of large sharks and tunas in the tropical Pacific since the onset of commercial fishing in the 1950s. • Meyers and Worm (2005) indicate a global depletion of large predatory fish communities of at least 90% over the past 50–100 years. The authors suggest that declines are “even higher for sensitive species such as sharks.” • Dulvy et al. (2008) state that “globally, three-quarters (16 of 21) of oceanic pelagic sharks and rays have an elevated risk of extinction due to overfishing.” • Graham et al. (2001) found an average decrease of 20% in the catch rate of sharks and rays off New South Wales, Australia, between 1976 and 1997.

95 Table 2. Marine mammal species in important conservation sites. “Irreplaceable areas” contain species found nowhere else. Figures from Pompa et al. (2011; supplt. material). 1Monachus schauinslandi, 2Arctocephalus galapagoensis 3A. philippii 4Inia geoffrensis, Trichechus inunguis (both freshwater) and Sotalia fluviatilis 5Monachus monachus 6Platanista minor (freshwater), 7Platanista gangetica (freshwater), 8Lipotes vexillifer (freshwater), 9Pusa sibirica (freshwater), 10Pusa caspica, 11Cephalorhynchus commersonii and A. gazella. *VU = Vulnerable, EN = Endangered, CR = Critically Endangered, LR Lesser Risk, EX Extinct, CE Critically Endangered; V = Vulnerable, RS = Relatively Stable or Intact. Data from Olson and Dinerstein (2002).

Key conservation Number Endemic/ Risk category Number and name of the Estimated sites of small- for each ecoregion* conservation species range ecoregion* status of the ecoregion* Highest richness South African 16 4 VU, EN 209: Benguela Current V 211: Agulhas Current RS Argentinean 15 4 VU, EN 205: Patagonian Southwest V Atlantic Australian 14 4 VU, EN 206: Southern Australian RS 222: Great Barrier RS Baja Californian 25 7 VU, EN, CR 214: Gulf of California CE Peruvian 19 5 VU, EN 210: Humboldt Current V Japanese 25 7 VU, EN, LR 217: Nansei Shoto CE New Zealand 13 2 VU, EN, LR 207: New Zealand V Northwestern 25 7 VU, EN, LR 216: Canary Current CE African Northeastern 25 7 VU, EN, LR 202: Chesapeake Bay V American Irreplaceable Hawaiian Islands 11 1 EN 227: Hawaiian Marine V Galapagos Islands 12 1 VU 215: Galapagos Marine V San Félix and Juan 13 1 VU 210: Humboldt Current V Fernández Islands Amazon River 24 1 VU 147: Amazon River/Flooded RS Forests Mediterranean Sea 15 1 CR 199: Mediterranean Sea CE Indus River 16 1 Not Listed Not Listed Not Listed Ganges River 17 1 EN Not Listed Not Listed Yang-tse River 18 1 EX 149: Yang-Tse River And CE Lakes Baikal Lake 19 1 LR 184: Lake Baikal V Caspian Sea 110 1 VU Not Listed Not Listed Kerguelen Islands 111 1 Not Listed Not Listed Not Listed

96 Figure 1. Geographic distribution of marine mammal species richness (left column) for A. Pinnipeds; B. Mysticetes; C. Odontocetes. Figure from Pompa et al.(2011).

97 Figure 2. At-sea distribution of threatened seabirds around the globe. Each polygon represents the range map for one threatened species. Areas of darkest blue show the areas of the ocean where the ranges of the greatest number of threatened species overlap. Figure from Birdlife International (2011).

Figure 3. Worldwide distribution of albatross and petrels. Figure from Birdlife International (2011).

Appendix – Appropriate management strategies

98 Appropriate management procedures may vary greatly between different fisheries, regulatory frameworks and species.To some extent, assessment of harvest control rules and other management strategies must therefore be addressed o a case-by-case basis. However, general guidelines for appropriate management are still relevant and useful. For fisheries managed using catch limits or TACs, these guidelines have been derived largely from the guidance provided for implementation of the Magnuson-Stevens Fishery Conservation and Management Act used for fishery management in the U.S. (Restrepo and Powers 1998; Restrepo et al. 1998). While other countries have different regulatory frameworks, similar strategies to those suggested in Restrepo et al. (1998) are used throughout the world where stock assessments are available and catch limits are employed (e.g., Australian Department of Agriculture, Fisheries and Forestry 2007; NZ Ministry of Fisheries 2008; DFO 2009). Commonly accepted strategies include setting fishing mortality rates safely below FMSY (or equivalent) to account for uncertainty; reducing when stocks fall below biomass target reference points (generally around BMSY or 40% of unfished biomass); and reducing fishing mortality when stock falls below a critical level where recruitment is impaired. Management reference points are assumed to be valid unless scientific information exists to suggest otherwise, e.g., scientific assessment or controversy that strongly suggests current reference points are not appropriate for the species under assessment.

In general, the minimal attributes of an appropriate management strategy include: 1. process for monitoring and conducting “assessments” (not necessarily formal stock assessments). Monitoring should occur regularly, though the frequency of assessments needed may vary depending on the variability of the stock and the extent of issues encountered in the management process. 2. Rules that control the intensity of fishing activity or otherwise ensure the protection of stock productivity. 3. process to modify rules according to assessment results, as needed. 4. process to gather appropriate types and extent of information needed to be able to review and revise management on timely basis 5. process to respond to knowledge gaps with appropriate levels of investment in research and development activities 6. A process to consult with all relevant stakeholders, secure their perceptions, collate their issues, and provide feedback and mechanisms for such information to be presented to, and incorporated as appropriate, into continuous improvement of the management system.

Some effective management strategies

For data-rich or data-moderate stocks that have quota-based management, “highly effective” management strategy is one that: • Incorporates an up-to-date scientific stock assessment that allows managers to determine if stocks are healthy and to set appropriate quotas; • Uses appropriate limit and target reference points for stock and fishing mortality; • Chooses risk-averse policies rather than risky, yield-maximizing policies; • Includes buffers in the TAC to account for uncertainty in stock assessments o Set Allowable Biological Catch (ABC) and Annual Catch Limit (ACL) at less than the Over- Fishing Level (OFL long term mean of MSY) to account for scientific uncertainty (survey data o stock size, etc. can reduce scientific uncertainty); o Set Total Allowable Catch (TAC) at less than AB to account for management uncertainty (monitoring catch, etc. can reduce management uncertainty);

99 o As a rule of thumb, TAC should have less than 30% p* (likelihood) of exceeding OFL; or TAC should be set such that is 25% below the threshold fishing pressure, e.g. Fmsy (Restrepo et al. 1998) o Stocks with low biomass, high vulnerability, and high uncertainty warrant greater protection against overfishing (e.g., more conservative harvest control rules/ greater buffers in setting TAC and/or closer monitoring of stocks). • Takes into account other sources of mortality (e.g., recreational fishery, bycatch of juveniles, etc.) and environmental factors that affect stock, such as oceanographic regime; • Incorporates strategy for maintaining or rebuilding stock productivity: o no-fishing point when biomass is below the limit reference point; o decrease in F when biomass is below the target reference point or is declining (whether declines are due to fishery or environmental factors). • Employs an effective strategy to prevent overcapitalization; • Has been demonstrated effective (e.g., stock productivity has been maintained over multiple generations), or if stock productivity has not been maintained or is declining, have adjusted management accordingly.

Effective management in data-poor fisheries (more information on data-poor evaluation methods below)

Whether managed stocks are data-rich or data-poor, management must include a strategy to ensure that stock productivity is maintained in order to be considered effective. This strategy should include a process for monitoring and conducting “assessments” of some kind (not necessarily formal stock assessments), rules that control the intensity of fishing activity or otherwise ensure the protection of a portion of the spawning stock, and system of adaptive management, such that rules are modified according to assessment results, as needed (Smith et al. 2009; Phipps et al. 2010).

There are some relatively reliable methods for setting catch limits in data-poor fisheries, including: An Index Method (AIM), which involves fitting a relationship between population abundance indices and catch; Depletion-Corrected Average Catch (DCAC), which allows managers to estimate a sustainable yield based on average catch over a set time period, adjusting for initial declines in abundance due to harvesting; and extrapolation methods, or relying o inferences from related or “sister” stocks, with the use of precautionary buffers in case the data-poor stocks are more vulnerable than the related data-rich stocks (Honey et al. 2010). Other techniques recommended for data-poor stocks include the use of productivity-susceptibility analysis (PSA) to highlight stocks that are particularly vulnerable to overexploitation (Patrick et al. 2009; Honey et al. 2010) and setting catch limits based on historical catch from a period of no declines, with targets set at 75% of average catch if biomass is believed to be healthy, 50% of average catch if biomass is expected to be below target levels but above the point where recruitment would be impaired, and 25% of average catch if the stock is depleted (Restrepo and Powers 1998).

Other than constraining fishing mortality (e.g., through TACs), fisheries may be credited for employing alternative strategies that are widely believed to be help maintain stock productivity. Some examples of effective alternative strategies are spatial management, including protecting large proportion of coastline in reserves and/or protecting known spawning aggregations with seasonal or spatial closures (e.g., Johannes 1998), or protecting females, which preserves the spawning per recruit of the population as long as fertilization does not decrease (e.g., Dungeness crab; Chaffee et al. 2010). Finally, stocks may be subject to low mortality in a data-poor fishery as result of low susceptibility, e.g., if the species is

100 small enough to fit between the mesh of the nets or is not attracted to the type of bait used (low susceptibility is generally more applicable as protection for non-target stocks).

Management of data-poor stocks – alternatives to MSY-based management

For data-poor stocks, management should: • Include a process for monitoring and assessment, such as recording trends in CPUE and size structure, or estimating FLEP, or comparison of abundance index to historical high (see glossary), unfished, or marine reserve levels: o Trends in CPUE are appropriate only if technology has not changed, there is no hyperstability, and abundance is shown to be proportional to CPUE; o Trends in size structure must also be monitored to avoid depletion of large individuals. • Include a strategy for protecting spawning stock, such as: o Estimate sustainable yield based on Depletion Corrected Average Catch (DCAC), An Index Method (AIM), or another accepted strategy; o Protect large portion of spawning stock in marine reserves (at least 50%, including important spawning areas if applicable) or close hotspots to fishing (for bycatch species); o Enforce size, sex, and/or season limitations that are likely to be effective in protecting spawning stock productivity (e.g., Dungeness crab 3-S management); o Extrapolate based o data-rich related or “sister” stocks, with precautionary buffers in place to account for potential differences in the stocks’ life histories; o Maintain exploitation rates at very low levels (e.g., experimental fishery) until more data can be collected, or o Base TAC on average historical catch during period of time with no declines in abundance (TAC should be set at no more than 75% of average catch if stock is believed to be healthy, 50% if believed to be below target levels, and 25% if believed to be overfished. Note: as there is generally n data to assess whether a stock is healthy, TAC should not be more than 50% of historical catch unless there is strong scientific reason to believe that stocks are above BMSY). • Allow for adaptive management so that fishing strategy is adjusted if assessment/monitoring indicates that stock is declining or below target levels; • Have been demonstrated effective (e.g., stock productivity has been maintained over multiple generations) or if stock productivity has not been maintained or is declining, management has been adjusted accordingly.

Procedures for monitoring/assessing stocks and procedures for protecting spawning stock must be in place, and be demonstrated effective, to qualify management strategy as “highly effective”. If measures are expected to be effective, e.g., through analogy with similar systems, but have not been demonstrated effective in this fishery, management is “moderately effective”. If measures are not expected to be effective, management strategy is “ineffective”.

Appropriate management also depends o the conservation concern associated with the stock. In addition to the precautionary elements listed above, stocks that are endangered or threatened also require a recovery plan and/or best management practices designed and demonstrated to reduce mortality and allow the stock to recover.Overfished and depleted stocks require a rebuilding plan.

Data–poor fishery evaluation methods

101 Sequential trend analysis (index indicators)

Sequential analysis comprises broad suite of techniques used to analyze time series data in order to detect trends in a variable (or in various indices) and infer changes in the stock or population. Sequential analyses can encompass wide range of data types and requirements (Honey et al. 2010). Examples include: DCAC, time series of catch statistics, survey/weight/length-based reference points, trophic indices, and spawning potential ratio (SPR) analogues (Honey et al. 2010).

Depletion-corrected average catch (DCAC) uses only catch time-series data supplemented with educated guesses for few supplementary parameters. Therefore, it is likely of practical use for many data-poor fisheries on long-lived species (e.g., natural mortality, 0.2) (Honey et al. 2010). The ability of this method to identify sustainable yields from simple data input makes DCAC useful as first-step estimate for an allowable catch level along with other data-poor methods. See: http://nft.nefsc.noaa.gov/ for the NOAA toolbox to perform DCAC analysis (Honey et al. 2010).

Vulnerability analysis

Productivity and susceptibility analysis of vulnerability – The Productivity-Susceptibility Analysis of vulnerability (PSA) method is used to assess a stock’s vulnerability to overfishing, based on relative scores derived from life-history characteristics. Productivity, which represents the potential for stock growth, is rated semi-quantitatively from low to high on the basis of the stock’s intrinsic rate of increase (r), von Bertalanffy growth coefficient (k), natural mortality rate (M), mean age at maturity, and other metrics (Patrick et al. 2009; Patrick et al. 2010; Field et al. 2010; Cope et al. 2011; Honey et al. 2010).

To assist regional fishery management councils in determining vulnerability, NMFS elected to use a modified version of a productivity and susceptibility analysis (PSA) because it can be based on qualitative data, has a history of use in other fisheries, and is recommended by several organizations as reasonable approach for evaluating risk (Patrick et al. 2010). Patrick et al. (2010) evaluated six U.S. fisheries targeting 162 stocks that exhibited varying degrees of productivity and susceptibility, and for which data quality varied. Patrick et al. (2010) found that PSA was capable of differentiating the vulnerability of stocks along the gradient of susceptibility and productivity indices. The PSA can be used as flexible tool capable of incorporating region-specific information on fishery and management activity. Similar work was conducted by Cope et al. (2011) who found that PSA is a simple and flexible approach to incorporating vulnerability measures into complex stock designations while also providing information helpful in prioritizing stock- and complex-specific management.

Extrapolation (Robin Hood Method)

When very limited or n data are available for a stock or specific species in a region, then managers may need to rely o extrapolation methods to inform decisions. Often, low-value stocks are data-poor (Honey et al. 2010). This method is termed the “Robin Hood” approach in Australia because it takes information and scientific understanding in data-rich fisheries and “gives” inferences to the data-poor fisheries (Smith et al. 2009). Data may include: (1) the local knowledge of the fishers and resource users; and/or (2) scientific research and ecosystem understanding from “sister” systems thought to be similar (Honey et al. 2010). Extrapolation from similar systems or related species may offer an informed starting point from which managers can build precautionary management (Honey et al. 2010). In these

102 situations, life-history characteristics, potentially sustainable harvest levels, spawning behavior, and other information can be gleaned from nearby stocks, systems, or related species (Honey et al. 2010).

Decision-making methods

Decision trees

Decision trees provide systematic, hierarchical frameworks for decision-making that can scale to any spatial, temporal, or management context in order to address a specific question. decision tree may be customized to meet any need (Honey et al. 2010). Trees may include: identification of reference points based o stock characteristics and vulnerability (Cope and Punt 2009); fostering of fine-scale, transparent, and local management (Prince 2010); and, estimation and refinement of an appropriate Total Allowable Catch (TAC) level (Wilson et al. 2010).

Management strategy evaluation

Management Strategy Evaluation (MSE) is a general modeling framework designed for the evaluation of performance of alternative management strategies for pursuing different objectives (Honey et al. 2010). This approach simulates the fishery’s response to different management strategies (e.g., different TAC levels, seasonal closures, or other effort reductions) (Honey et al. 2010). Assuming sufficient quality data exist, MSE may be useful for assessing the effectiveness of different policy options (Honey et al. 2010).

In addition, a study by Dowling et al. (2008) developed harvest strategies for data-poor fisheries in Australia. Strategies included: (i) the development of sets of triggers with conservative response levels, with progressively higher data and analysis requirements at higher response levels, (ii) identification of data gathering protocols and subsequent simple analyses to better assess the fishery, (iii) the archiving of biological data for possible future analysis, and (iv) the use of spatial management, either as the main aspect of the harvest strategy or together with other measures (Honey et al. 2010).

Cooperative research an co-management to overcome data-poor situations

recent study by Fujita et al. (2010) identified opportunities for cooperative research and co- management that would complement (but not replace) existing top-down fishery regulations. They conclude that management and data collection would improve for some small-scale fisheries if they started: collecting data at the appropriate spatial scales; collecting local information, improving the quality of data, and overcoming constraints o data; providing ecosystem insight from a small/local scale for new and different perspectives; reducing conflicts among fishermen, scientists, and regulators; and improving the responsiveness of fisheries management to local needs. Fujita et al. (2010) suggest that scientists and managers should further develop cooperative strategies (e.g., cooperative research and co-management) and include them in the management framework.

Effective management of fishery on a non-native species Effective management of fishery for non-native species may include: • Mitigation strategies aimed at eradication, reversing establishment, or maintenance at low abundance, as deemed appropriate and feasible for that particular case, • Adaptation strategies that allow for recovery of species impacted by the non-native species, • Containment measures such as fishing at the boundaries of the stock to prevent further spread, and/or

103 • Provisions to prohibit further introductions of any other alien species.

Appendix 5 – Bycatch reduction approaches

In general, fisheries should address bycatch with the following approaches: • Monitor bycatch rates (using adequate observer coverage), • Have some scientific assessment of impacts on bycatch populations • Incorporate strategies that assure bycatch is minimized, such as: o Enforcing effective and appropriate bycatch caps, o Closing hotspots or implementing seasonal closures, o Promoting effective gear modifications such as BRDs, TEDs, etc. o Adopting bycatch-reducing strategies such as night setting, o Using the best available management techniques that have been demonstrated in this or a similar system to effectively constrain bycatch rates.

The effectiveness of various bycatch reduction approaches is synthesized from primary literature and reviewed below. To be considered “highly effective”, all required measures and at least one primary measures should be in place.

Seabird sources are Løkkeborg (2008) (general conclusions and Table 3, including percent effectiveness of some modification/region strata) and SBWG 2010 (Annexes 3–8). *Secondary measures may be useful in conjunction with primary measures. Turtle sources are FAO 2009 (Tables 1 and page 79) and Gilman and Lundin (2008) (Table 3). Shrimp trawl modifications sources are Eayers 2007 and Gillet 2008 (Box 14). Sharks and marine mammals from Gilman and Lundin (2008) (Table 3). General information on fishing technologies can be found at http://www.fao.org/fishery/en and a list of bycatch reduction literature can be found here: http://www.bycatch.org/articles.

Primary/ Gear/taxon/modification secondary Effectiveness/notes measure* General strategies (good for all gears/taxa) Considerable difference between experimental and real-world effectiveness. “Three common themes to successful implementation Monitoring and Require- of bycatch reduction measures are long-standing collaborations compliance ment among the fishing industry, scientists, and resource managers; pre- and post-implementation monitoring; and compliance via enforcement and incentives” (Cox, Lewison et al. 2007). Area/time closures. Generally very effective, though more so when based on data such as tagging or bycatch data. Perhaps only a secondary mitigation measure for birds (Løkkeborg 2008). Avoid bycatch hotspots Primary Alternatively, move when interaction rates are high. Effective for all fisheries, especially with fleet communication. Closures for one taxon without commensurate reduction in effort can increase bycatch of other taxa. Bycatch caps Primary I.e., fishery closes when cap exceeded.

104 Not effective. “We conclude that, overall, CMMB has little potential for benefit and a substantial potential for harm if implemented to solve most fisheries bycatch problems. In particular, CMMB is likely to be effective only when applied to short-lived and highly fecund species (not the characteristics of most bycatch-impacted species) Bycatch fees, and to fisheries that take few non-target species, and especially few Compensatory Mitigation Secondary, non-seabird species (not the characteristics of most fisheries). Thus, Strategies for Marine at best CMMB appears to have limited application and should only be Bycatch (CMSMB) implemented after rigorous appraisal on a case-specific basis; otherwise it has the potential to accelerate declines of marine species currently threatened by fisheries bycatch” (Finkelstein, Bakker et al. 2008). May be useful, but only as a complementary measure (Žydelis, Wallace et al. 2009). Pelagic longline No single solution to avoid incidental mortality of seabirds in pelagic Seabirds (albatrosses and longline fisheries. Most effective approach is streamer lines combined Best petrels) with branchline weighting and night setting. Best practices are followed for line setting and hauling (e.g., SRWG 2010). Proven effective in Southern Hemisphere. Streamer lines and Night setting Primary weighted lines should also be used when interacting with nocturnal birds/fishing during bright moon. Proven to be effective in North Atlantic. Should be paired and/or Streamer/scarer lines Primary weighted lines in North Pacific. Paired lines need more testing. Light configuration not recommended. Weighted branch lines Primary Must be combined with other measures. Offal discharge Secondary Not yet established but is thought to assist. management Insufficiently researched; there have been operational difficulties on some vessels. Effective in Hawaii in conjunction with bird curtain and Sidesetting Secondary weighted branch lines. Japanese research conclusions must be combined with other measures. Untested in Southern Hemisphere. Line shooter and mainline tension, bait caster, live - Not recommended. bait, thawing bait Underwater setting chute, Insufficient research. Blue-dyed bait may be only effective with squid hook design, olfactory - bait. Results inconsistent across studies. deterrents, blue-dyed bait Turtles Replacement of J and tuna Primary Wide circle hook with

105 Bait change Primary Fish instead of squid. Prohibit wire leaders Primary Deeper setting Primary Avoid surface waters. Shark repellants - Insufficient research. Circle hooks Marine mammals Weak hooks, deterrents, - Insufficient research. echolocation disruption Other finfish (including juvenile targets) Circle hooks May help reduce mortality of billfish and tunas. Shellfish Not problematic Bottom longline (Many measures similar to pelagic longline) No single solution to avoid incidental mortality of seabirds in demersal longline fisheries. No combination specified: assume streamers, weighted and night setting, or Chilean longline method Seabirds (albatrosses and Best (vertical line with very fast sink rates—considered effective even petrels) without other measures; widely used in South American waters and SW Atlantic). Best practices are followed for line setting and hauling (e.g., SRWG 2010). Effective, but must be used properly (streamers are positioned over Streamer/scarer lines Primary sinking hooks). Better when combined with, e.g., night setting, weighting, or offal control. Must be combined with other measures, especially streamers, offal Weighted lines Primary control and/or night setting. Night setting Primary Same as pelagic. Haul curtain (reduce bird Can be effective, but must use strategically as some birds become access when line is being Secondary habituated. Must be used with other measures. hauled) Offal discharge control Must be used in a combo, e.g., with streamers, weighting, or night (discharge homogenized Secondary setting. offal at time of setting) Insufficiently researched; there have been operational difficulties on Side setting Secondary some vessels. Hook design, olfactory deterrents, underwater Insufficiently researched. Blue-dyed bait, thawed bait, and use of line setting chutes, blue-dyed - setter not relevant in demersal gear. bait, thawed bait, use of line setter Turtles, sharks, mammals, other finfish, shellfish See pelagic longlines Trawl

106 Little work has been done on seabird bycatch mitigation in trawl fisheries (pelagic and demersal). There is no single solution to avoid incidental mortality of seabirds in trawl fisheries. The most effective Seabirds (albatrosses and approach is offal discharge and discards control, through full retention Best petrels) of all waste or mealing (the conversion of waste into fish meal reducing discharge into sump water) plus streamer lines. Effectiveness of other offal control measures such as mincing and batching is not clear. Minimum requirement Limited waste control No discharge of offal or discards during shooting and hauling. for + modifier Reduce cable strike Scarers recommended even when offal/discard management is in through bird scaring wires Primary place. Snatch block recommended on theory. or snatch block Reduce net entanglement through net binding, net Recommended on theory. weights, net cleaning Net jackets - Not recommended. Reduced mesh size, acoustic scarers, warp - Effectiveness not yet established. scarers, bird bafflers, cones on warp cables Turtles Any modification to the trawl to reduce the capture of turtles, principally in tropical/subtropical shrimp trawls. Typically a grid or large-hole mesh designed to prevent turtles from entering the codend. The only designs approved for use in the US warm-water shrimp fisheries are hard TEDs (i.e.,“hooped hard TEDs” such as NMFS, Coulon and Cameron TEDs, “single grid hard TEDs” such as the Turtle excluder device Primary Matagorda, Georgia, or Super Shooter TED, and the Weedless TED) (TED) and the Parker Soft TED (the latter only in offshore and inshore waters in Georgia and South Carolina). Hard TEDs that are not approved for use in the shrimp fisheries are used in the Atlantic summer flounder bottom trawl fishery. TEDs must be used in conjunction with escape hatches, which also vary in size and design. More details on TED/hatch designs and US regulations can be found in Eayers (2007). Sharks TEDs generally allow large animals to escape, e.g., sharks (Belcher and TED Jennings 2010). Highly variable depending on net type and TED used. BRD made little difference (fisheye). Marine mammals TED/BRD Grids generally allow large animals to escape. Other finfish

107 A BRD is any modification designed principally to exclude fish bycatch from shrimp trawls. Catch separator designs include hard grids (e.g., Nordmore grid) and soft mesh panels attached at an angle inside the Bycatch reduction device trawl net as well as the Juvenile and Trash Excluder Device (JTED), (BRD): Catch separators which has a grid/mesh design partially covering the inside of the trawl net. Hard grids are generally seen as more effective than soft panels. Effectiveness of JTED unknown. Designed for strong-swimming fish to actively escape (shrimp are more passive swimmers). Most are located in the codend (e.g., BRD: Active swimmer fisheye and fishbox) although others can be in the body of the trawl escape hatches (square mesh window, composite square mesh panel, radial escape section). BRD: Square-mesh codend Square mesh stays open under tow (unlike diamond mesh). E.g., the cone. Stimulates fish to swim forward through escape BRD assist hatches like the fisheye, square mesh window or radial escape section. Inclusion of increased mesh sizes in the upper wings and upper netting panel immediately behind the headrope crown, coupled with Coverless trawl reduced headline height, encourages the escape of fish species such as haddock and whiting in and around the mouth of the trawl. Triangular/diamond-shaped cut in the top of the codend (e.g., flapper), changes to ground chain settings, headline height reduction, Rigging modification a length of twine stretched between the otter boards to frighten fish, large mesh barrier across trawl mouth and large cuts in the top panel of the net ahead of the codend. Semi-pelagic rigging Avoid contact with seabed. Trawl separator (Rhule Reduces cod catch in haddock trawls by separating catch and trawl) releasing cod from the net. Shellfish TEDs generally allow large animals to escape (jellyfish). Downward TED facing TEDs may also allow benthic invertebrates to escape. BRD e.g., Nordmore grid Effective for jellyfish? Longer sweeps between the otter board and trawl can reduce Rigging modification invertebrate bycatch. Semi-pelagic rigging Avoid contact with seabed. Other BRD Seahorses, sea snakes in Australian prawn fisheries. Gillnet Seabirds Less research than for trawls. Visual and acoustic alerts - Pingers may also reduce seabird bycatch (1 study in Lokkeborg 2008). Turtles Reduces entanglement as the net is stiffer. Good for both demersal Use lower profile nets Primary and drift nets. Creates slack in the net, increasing chances of entanglement (rather Use of tie-down ropes Negative than gilling). Set nets perpendicular to - Insufficient research. May reduce interactions with nesting females. shore Use deterrents - Insufficient research. Pingers, shark silhouettes, lights or chemicals. Deep setting - Insufficient research. Avoid upper water column (above 40m).

108 Sharks Unknown Marine mammals Acoustic devices to keep cetaceans (and possibly pinnipeds) away from nets. Effectiveness appears to vary considerably depending on Pingers fishery and cetacean species: http://cetaceanbycatch.org/pingers_effectiveness.cfm Shellfish Weak buoy lines Mesh size Purse Seine Seabirds Not problematic Turtles Avoid encircling turtles. Restrict setting on FADs, logs and other Avoid turtles Primary debris. Use of modified FAD - Insufficient research. designs Sharks Avoid restrict setting on FADs, logs, other debris and whales. Avoid Avoid sharks Primary hotspots. Shark repellants - For deployment on FADs. Insufficient research. Marine mammals Backdown maneuver, Medina panel, deploy Primary rescuers Avoid mammals Restrict setting on mammals. Other finfish Sorting grids - Insufficient research. Avoid finfish Restrict setting on FADs. Shellfish Not problematic Pots and traps Turtles E.g., Diamondback terrapins in Floridian blue crab pot fishery (Butler BRDs Primary and Heinrich 2005). Marine mammals Weak lines Primary E.g., northern right whales, NE lobster fishery. Finfish, invertebrates BRDs Primary

Appendix 6 – Impact of fishing gear on the substrate

In order to assess fisheries for habitat impacts under the Seafood Watch ® criteria, we developed a matrix to help determine the potential impacts that different fishing gear may have on various habitat

109 types. The matrix was developed based on similar work done by the New England Fisheries Management Council (NEFMC 2010) and the Pacific Fisheries Management Council (PFMC 2005).

The NEFMC (2010) created “Swept Area Seabed (SASI) model” that assessed habitat susceptibility and recovery information. Susceptibility and recovery were scored (0–3) based on information found in the scientific literature and supplemented with professional judgment when research results were deficient or inconsistent.

“Vulnerability was defined as the combination of how susceptible the feature is to a gear effect and how quickly it can recover following the fishing impact. Susceptibility was defined as the percentage change in functional value of a habitat component due to a gear effect, and recovery was defined as the time in years that would be required for the functional value of that unit of habitat to be restored (ASFMC 2010).”

The PFMC (2005) created similar habitat sensitivity scale (0–3) that represents the relative sensitivity of different habitats to different gear impacts. The sensitivity of habitats from the PFMC (2005) was based o actual impacts reported in the scientific literature.

The relative impacts by gear and habitat type used for the Seafood Watch® matrix were based on the sum of sensitivity and recovery values from tables developed by the NEFMC (2010) (substrates) and the PFMC (2005) (biogenic). The NEFMC (2010) excluded deep-sea corals with extreme recovery times. The values for deep-sea corals in this matrix are the sum of the sensitivity and recovery scores from PFMC (2005). Other biogenic habitats that were not included in the NEFMC (2010) data tables include: seagrass, sponge reefs (rather than individual sponges) and maerl beds. Due to the slow recovery and importance of these habitat types, they have been given the same value as coral and sponge habitats, all of which are listed as “biogenic”.

Hall-Spencer and Moore (2000) examined the effects of fishing disturbance on maerl beds. Maerl beds are composed of calcareous alga and form complex habitats with a high degree complexity. The associated species assemblages have high diversity (Hall-Spencer and Moore, 2000). Hall-Spencer and Moore (2000) showed that four years after an initial scallop-dredging disturbance had occurred, some fauna, such as the bivalve Limaria hians, had still not re-colonized the trawl tracks. Similarly, work by Sainsbury et al. (1998; in Kaiser et al. 2001) suggests that recovery rates may exceed fifteen years for sponge and coral habitats off the western coast of Australia.

Hydraulic clam dredges are rated as high concern according to Seafood Watch ®. There are very few studies on the impact of this gear type, so we have relied on expert opinion (NEFMC 2010). Hydraulic clam dredges are used primarily in sand and granule-pebble substrates because they cannot be operated in mud or in rocky habitats (NEFMC 2010). This gear type is effective at pulverizing and/or removing solids and flattening out seafloor topography (NEFMC 2010). In addition, the habitats where this gear type is used are very susceptible to hydraulic dredges; recovery is moderate on average (NEFMC 2010). This leads Seafood Watch® to rate hydraulic dredges as “high concern .” Hydraulic dredges do not operate o deep-sea coral or other biogenic habitats.

Neckles et al. (2005) found significant differences in eelgrass biomass between disturbed and reference sites up to seven years after dragging. The authors projected that it would require a mean of 10.6 years for eelgrass shoot density to recover in areas of intense dragging.

110 Demersal seines were not evaluated in the reports by Fuller et al. (2008), Chuenpagdee et al. (2003), NEFMC (2010) or PFMC (2005). Demersal seines include: Danish seines, Scottish fly-dragging seines and pair seines. These seines are similar to some bottom trawl gear in that they have a funnel shaped net with a groundrope. They are generally hauled by wires or ropes, and although they are lighter than some bottom trawl gear, they create habitat disturbance (Rose et al. 2000; Thrush et al. 1998; Valdemarason and Suuronen 2001). A review of trawling impacts by Jones (1992) grouped bottom trawling, dredges and Danish seines together as having similar impacts o the benthos when assessing the environmental effects of bottom trawling. However, studies have demonstrated Danish seines to have less impact o the substrate compared to bottom trawls (Gillet 2008). Therefore, in our matrix they are given an intermediate score as more damaging that bottom longlines and bottom gillnets, but less damaging than bottom trawls. Beam trawls were also not included in the reports, but were considered to be similar to otter trawls.

The matrix developed from the sources referenced above is shown o the next page. For use in evaluating the Fisheries Criteria, these data have been summarized into categories (low impact, moderate, moderate-severe, severe, and very severe) to simplify use of the table.

111 Habitat impacts matrix: Relative impacts by gear and habitat type.

Mud Sand Granule-pebble Cobble Boulder Deep-sea corals ** low high low high low high low high low high Line, Vertical (BL/2) 0.5 0.5 0.6 0.5 0.8 0.8 1.0 0.9 1.0 1.0 1.3 Longline, Bottom**** 0.7 0.7 0.9 0.8 1.4 1.3 1.6 1.5 1.7 1.7 2.0 Trap (lobster and deep-sea red crab) 1.3 1.3 1.2 1.2 1.8 1.7 2.0 1.9 2.1 2.1 1.3 Gillnet, Bottom**** 1.3 1.3 1.5 1.4 2.0 1.9 2.2 2.1 2.3 2.3 3.0 Bottom Longline, Gillnet 1.0 1.0 1.2 1.1 1.7 1.6 1.9 1.8 2.0 2.0 2.5 Seine, Bottom (BL,G+TBO/2) 1.8 1.7 2.0 1.9 2.5 2.3 2.7 2.5 2.6 2.6 3.6 Trawl, Shrimp (BS+TBO/2) 2.2 2.1 2.5 2.3 3.0 2.6 3.0 2.8 3.0 2.9 4.1 Trawl, Bottom Otter 2.6 2.4 2.9 2.7 3.4 2.9 3.4 3.1 3.3 3.2 4.6 Dredge, New Bedford Scallop 2.6 2.4 3.0 2.8 3.5 3.0 3.5 3.2 3.3 3.2 5.1 Dredge, Hydraulic Clam n/a*** 4.4 4.0 4.9 4.5 n/a*** Explosives/Cyanide 6 6 6 6 6 6 6 6 6 6 6

*Shrimp trawls tend to be lighter than bottom otter trawls for fish and do notneed to touch theseabed to beeffective. ** Most biogenic habitats (macroalgae, cerianthid anemones, polychaetes, sea pens, sponges, mussel and oyster beds) areincorporated into thescores for each substrate/gear combination in thetable. NEFMC 2010 specifially excluded deep-sea corals. Thenumbers for deep-sea corals in this matrix are the sum of the sensitivity and (standardized) recovery scores in PFMC 2005. Other biogenic habitats that werenot included in theNEFMC data tables include seagrass meadows, sponge reefs (rather than individual sponges) and maerl beds. Use the 'deep-sea corals' column for thesehabitats. *** Scores not determined for hydraulic dredges in these habitats as the gear is assumed to not operate in them (NEFMC 2010). **** NEFMC 2010 groups bottom longlines and gillnets as 'fixed gear' (not shown in table). Thesescores have been disaggregated here for substrate habitats only by adding 0.4 to the aggregated score for gillnets and subtracting 0.4 for longlines, basedon therelativeimpacts shown in PFMC 2005 (i.e. that gillnets are generally more damaging than longlines).

The values above are the sum of sensitivity and recovery values in tables from Section 5.2 in Part of (NEFMC 2010) (substrates) and Tables 4 and in Appendix C, Part of (PFMC 2005) (biogenic). Gear types in black are from the Swept Area Seabed Impact (SASI) model used for the NEFMC EFH process (NEFMC 2010). Gear types in red are derived from those in black. Substrate types are self-explanatory except that mud includes clay-silt and muddy sand, and boulder includes rock. The energy regime is used here as a proxy for natural disturbance, with a cutoff between low and high stability at 60m depth. Most biogenic habitats (macroalgae, cerianthid anemones, polychaetes, sea pens, sponges, mussel and oyster beds) are incorporated into the scores for each substrate/gear combination in the table. NEFMC (2010) specifically excluded deep-sea corals with extreme recovery times. The numbers for deep-sea corals in this matrix are the sum of the sensitivity and (standardized) recovery scores from PFMC (2005). Other biogenic habitats that were not included in the NEFMC data tables include seagrass meadows, sponge reefs (rather than individual sponges) and maerl beds. Use the “deep-sea corals” column for these habitats

112 Appendix 7 – Gear modification table for bottom tending gears

Spatial protection

Reducing the footprint of fishing through spatial management can be one of the most effective ways to mitigate the ecological impact of fishing with habitat-damaging gears (Lindholm et al. 2001; Fujioka 2006). The relationship between gear impacts, the spatial footprint of fishing and fishing effort (i.e., frequency of impact) is complex (Fujioka 2006) and cannot be quantified precisely in Seafood Watch® assessments. Nevertheless, criteria should acknowledge the benefits of conservative habitat protection efforts by adjusting the habitat score. Thresholds for adjusting the habitat score due to habitat protection from the gear-type used in the fishery (50% protected to qualify as “strong mitigation” and 20% protected to qualify as “moderate mitigation”) are based on recommendations for spatial management found in the scientific literature as noted in Auster (2001). Auster recommends use of the precautionary principle when a threshold level of 50% of the habitat management area is impacted by fishing, with a minimum of 20% of regions in representative assemblages and landscape features protected in MPAs in order to minimize impacts o vulnerable species and sensitive habitats.

The table below gives examples of gear modifications that are believed to be moderately effective at reducing habitat impacts based on scientific studies. This table will be continually revised as new scientific studies become available. The main sources for the current table are He (2007) and Valdemarsen, Jorgensen et al. (2007).

Gear Modification Otter Semi-pelagic trawl rigging (trawl doors, sweeps and bridles off the bottom, also includes Trawls modifications such as short bridles and sweeps—most commonly used for shrimp, nephrops and other species that are not herded by sand clouds and bridles due to poor swimming ability) Quasi-pelagic trawl rigging/sweepless trawls (trawl doors remain in contact with the seafloor, remaining gear largely off the bottom, e.g., whiting in New England, flatfish in Alaska, red snapper in Australia) Lighter ground gear (e.g., fewer bobbins) Use of rollers instead of rockhoppers Trawl door modifications such as high aspect (smaller footprint), cambered (generally for fuel efficiency) or soft doors (e.g., self-spreading ground gear)

113 Appendix References

Appendix 1 Botsford, L. W., and A. M. Parma. 2005. Uncertainty in Marine Management. Pages 375-392 in E. A. Norse and L. B. Crowder, editors. Marine Conservation Biology: The Science of Maintaining the Sea's Biodiversity. Island Press, Washington, DC.

Clark, W.G. 1991. Groundfish exploitation rates based o life history parameters. Canadian Journal of Fisheries and Aquatic Sciences 48: 734-750.

Clark W. G. 2002. F35% revisited ten years later North American Journal of Fisheries Management 22:251-257.

Froese, R., T.A. Branch, A. Proelß, M. Quaas, K. Sainsbury, and C. Zimmermann. 2010. Generic harvest control rules for European fisheries. Fish and Fisheries: doi:10.1111/j.1467-2979.2010.00387.x

Goodman, D., M. Mangel, G. Parkes, T. Quinn, V. Restrepo, T. Smith and K. Stokes. 2002. Scientific Review of The Harvest Strategy Currently Used in The Bsai and Goa Groundfish Fishery Management Plans, North Pacific Fishery Management Council, Anchorage, AK. 15 p. Available at: http://www.fakr.noaa.gov/npfmc/misc_pub/f40review1102.pdf

Honey, K.T., J.H. Moxley and R.M. Fujita. 2010. From rags to fishes: data-poor methods for fishery managers. Managing Data-Poor Fisheries: Case Studies, Models & Solutions 1: 159-184.

ICES 2010. General context of ICES advice. Available at http://www.ices.dk/committe/acom/comwork/report/2010/2010/Introduction%20for%20Advice.pdf

Kell, L. T., Pastoors, M. A., Scott, R. D., Smith, M. T., Van Beek, F. A., O’Brien, C. M., and Pilling, G. M. 2005. Evaluation of multiple management objectives for Northeast Atlantic flatfish stocks: sustainability vs. stability of yield. ICES Journal of Marine Science 62: 1104-1117.

Myers R. A., Bowen K. G., and Barrowman N. J. 1999. Maximum reproductive rate of fish at low population sizes. Canadian Journal of Fisheries and Aquatic Sciences 56:2404-2419

New Zealand Ministry of Fisheries. 2008. Harvest Strategy Standard for New Zealand Fisheries. 25 p. Available at: http://fs.fish.govt.nz/Doc/16543/harveststrategyfinal.pdf.ashx

O’Farrell, M.R. and L.W. Botsford. 2006. Estimation of change in lifetime egg production from length frequency data. Canadian Journal of Fisheries and Aquatic Sciences 62: 1626-1639.

114 Phipps, K.E., R. Fujita, and T. Barnes. 2010. From paper to practice: incorporating new data and stock assessment methods into California fishery management. Managing Data-Poor Fisheries: Case Studies, Models & Solutions 1: 159-184.

Roughgarden, J. and F. Smith. 1996. Why fisheries collapse and what to d about it. Proc. Nat. Acad. Sci., (USA), 93:5078-5083

Appendix 2 Marine Stewardship Council (MSC). 2012. MSC Certification Requirements, version 1.2. Marine Stewardship Council, London, UK. 301 p. Available at: http://www.msc.org/documents/scheme- documents/msc-scheme-requirements/msc-certification-requirements-v1.2/view

Appendix 3 Abraham, E., J. Pierre, et al. 2009. Effectiveness of fish waste management strategies in reducing seabird attendance at a trawl vessel. 95 (3): 210-219

Alverson, D.L.; Freeberg, M.H.; Pope, J.G.; Murawski, S.A. A global assessment of fisheries bycatch and discards. FAO Fisheries Technical Paper. No. 339. Rome, FAO. 1994. 233p.

Baird, S.J., E. Bradford. 2000. Factors that may have influenced the capture of NZ fur seals (Arctocephalus forsteri) in the west coast South Island hoki fishery, 1991–1998. NIWA Technical Report 92. 3 p.

Baum, J.K. & Myers, R.A. (2004). Shifting baselines and the decline of pelagic sharks in the Gulf of Mexico. Ecol. Lett., 7, 135–145.

Baum, J.K., Myers, R.A., Kehler, D.G., Worm, B., Harley, S.J. & Doherty, P.A. (2003). Collapse and conservation of shark populations in the Northwest Atlantic. Science, 299, 389–392.

Birdlife International. 2011. http://www.birdlife.org/action/science/species/seabirds/index.html

BirdLife International 2010. Marine Important Bird Areas - Priority for the Conservation of Biodiversity. Cambridge, UK: BirdLife International.. ISBN 978-0-946888-74- 0. http://www.birdlife.org/community/wp-content/uploads/2010/10/marineIBAs.pdf.

Camhi, M.D., Valenti, S.V., Fordham, S.V., Fowler, S.L. and Gibson, C. 2009. The Conservation Status of Pelagic Sharks and Rays: Report of the IUCN Shark Specialist Group Pelagic Shark Red List Workshop. IUCN Species SurvivalCommission Shark Specialist Group. Newbury, UK. x + 78p.

Chuenpagdee, R., L. E. Morgan, et al. (2003). "Shifting gears: Assessing collateral impacts of fishing methods in US waters." Frontiers in Ecology and Environment 1(10): 517-524.

Chapple, T.K., Jorgensen, S.J., Anderson, S.D., Kanive, P.E., Klimley, A.P., Botsford, L.W. and Block, B.A. (2011). A first estimate of white shark, Carcharadon carcharias abundance off Central California. Biol. Lett. 00, 1-3.

Dufresne, S.P., Grant, A., Norden, W.S., Pierre, J. 2007. Factors affecting cetacean bycatch in a New Zealand trawl fishery. DOC Research and Development series 282. 18pp.

115 Dulvy, N.K., Baum, J.K., Clarke, S., Compagno, L.J.V., Corte´s, E., Domingo, A., et al. (2008). You can swim but you can_t hide: the global status and conservation of oceanic pelagic sharks and rays. Aquat. Conserv., 18, 459–482.

Ferretti, F., Myers, R.A., Serena, F. & Lotze, H.K. (2008). Loss of large predatory sharks from the Mediterranean Sea. Conserv. Biol., 22, 952–964. Fertl, D.; Leatherwood, S. 1997: Cetacean interactions with trawls: preliminary review. Journal of Northwest Atlantic Fishery Science 22: 219–248.

Fuller, S.D., C Picco, et al. 2008. How we fish matters: Addressing the ecologicalimpacts of Canadian fishing gear. Ecology Action Centre, Living Oceans Society, and Marine Conservation Biology Institute. 25pp.

Icelandic Ministry of Fisheries and Agriculture Fishing Gear-Danish Seine. Accessed on January 20, 2011. http://www.fisheries.is/fisheries/fishing-gear/danish-seine/

Morizur, Y.; Berrow, S.D.; Tregenza, N.J.C.; Couperus, A.S.; Pouvreau, S. 1999: Incidental catches of marine mammals in pelagic trawl fisheries of the northeast Atlantic. Fisheries Research 41: 297–307.

Myers, R.A., Baum, J.K., Shepherd, T., Powers, S.P. & Peterson, C.H. (2007). Cascading effects of the loss of apex predatory sharks from a coastal ocean. Science, 315, 1846–1850.

Myers, R.A. & Worm, B. (2005). Extinction, survival or recovery of large predatory fishes. Phil. Trans. R. Soc. Lond. B, 360, 13–20. Graham, K.J., Andrew, N.L. & Hodgson, K.E. (2001). Changes in relative abundance of sharks and rays on Australian south east fishery trawl grounds after twenty years of fishing. Mar. Freshwater Res., 52, 549– 61.

Mapping the World’s Sea Turtles. 2011. http://www.gearthblog.com/blog/archives/2011/07/mapping_the_worlds_sea_turtles.html?utm_sourc e=feedburner&utm_medium=feed&utm_campaign=Feed%3A+GoogleEarthBlog+%28Google+Earth+Blo g%29

Palsson, O.K. 2003. A length-based analysis of haddock discards in Icelandic Fisheries. Fisheries Research. 59: 437-446.

Pompa, S., Ehrlich,P.P.,Ceballos, G. 2011. Global distribution and conservation of marine mammals. PNAS. www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1101525108/- /DCSupplemental.

Rowe, S.J. 2007. review of methodologies for mitigating incidental catch of protected marine mammals. DOC Research & Development Series 283. 48pp.

Sullivan, B.J., P. Brickle, T.A. Reid, D.G. Bone and D.A.J. Middleton. 2006. Mitigation of seabird mortality on factory trawlers: trials of three devices to reduce warp cable strikes, Polar Biology 29 (2006), pp. 745–753.

116 Wallace, B.P. and others. 2010. Global patterns of marine turtle bycatch. Conservation Letters. 1-21

Watling, L. 2005. The global destruction of bottom habitats by mobile fishing gear. In: Marine Conservation Biology, The science of maintaining the sea’s biodiversity. Ed by E. Norse and L Crowder. pp198-210. Island Press.

Ward, P. & Myers, R.A. (2005). Shift in open-ocean fish communities coinciding with the commencement of commercial fishing. Ecology, 86, 835–847.

Weimerskirch, H, D. Capdeville and G. Duhamel. 2000. Factors affecting the number and mortality of seabirds attending trawlers and longliners in the Kerguelen area, Polar Biology. 23 pp. 236–249

Wieneke, J. and G. Robertson. 2002. Seabird and seal—fisheries interactions in the Australian Patagonian toothfish, Dissostichus eleginoides trawl fishery, Fisheries Research 54 (2002). pp. 253–265

SWOT (State of the World’s Se Turtles). 2011. http://seamap.env.duke.edu/swot

Appendix 4 Cope, J. 2011. Personal Communication. National Marine Fisheries Service, Northwest Fisheries Science Center.

Cope, J. and coauthors. 2011. A approach to defining stock complexes for U.S. west coast groundfishes using vulnerabilities and ecological distribution. In press, North Atlantic Journal of Fisheries Management

Cope, J. M. and A. E. Punt. 2009. Length-based reference points for data-limited situations: Applications and restrictions. Marine and Coastal Fisheries: Dynamics, Management, and Ecosystem Science. 1:169- 186.

Dick, E.J and A.D. MacCall. 2010. Estimates of sustainable yield for 50 data-poor stocks in the Pacific coast groundfish fishery management plan. NOAA-TM-NMFS-SWFSC-460. Available at : http://swfsc.noaa.gov/publications/TM/SWFSC/NOAA-TM-NMFS-SWFSC-460.pdf

Dowling, N.A., and others. 2008. Developing harvest strategies for low-value and data-poor fisheries: Case studies from three Australian fisheries. Fisheries Research 94 (2008) 380–390

Field, J., J. Cope and M. Key. 2010. A descriptive example of applying vulnerability evaluation criteria to California nearshore species. Managing data-poor fisheries: Case studies, models and solutions. 1:235- 246

Fujita, R. M., K. T. Honey, A. Morris, H. Russell and J. Wilson. 2010. Cooperative strategies in fisheries management: Transgressing the myth and delusion of appropriate scale. Bulletin of Marine Science. 86:251-271.

Honey, K.T. Moxeley, J.H., and Fujita, R. M. 2010. From rages to fishes. Managing Data-Poor Fisheries: Case Studies, Models & Solutions. California Sea Grant College Program. (1):159–184. ISBN number 978-1-888691-23-8

117 Johannes, R. E. 1998. The case for data-less marine resource management: Examples from tropical nearshore fisheries. Trends in Ecology and Evolution 13: 243–246.

MacCall, A. D. 2009. Depletion-corrected average catch: a simple formula for estimating sustainable yields in data-poor situations. – ICES Journal of Marine Science. 66: 2267–2271.

NMFS. 2011. Assessment Methods for Data-Poor Stocks Report of the Review Panel Meeting National Marine Fisheries Service (NMFS) Southwest Fisheries Science Center (SWFSC) Santa Cruz, California April 25-29, 2011. Available at: http://www.pcouncil.org/wp- content/uploads/E2a_ATT6_DATAPOOR_RVW_JUN2011BB.pdf

Patrick, W. and coauthors. 2010. Using productivity and susceptibility indices to assess the vulnerability of United States fish stocks to overfishing. Fish. Bull. 108:305–322.

Prince, J. D. 2010. Managing data poor fisheries: Solutions from around the world. Managing data-poor fisheries: Case studies, models and solutions. 1:3-20

Restrepo, V.R. and J. E. Powers. 1998. Precautionary control rules in US fisheries management: specification and performance. ICES Journal of Marine Science 56: 846–852.

Smith, D. et al. 2009. Reconciling approaches to the assessment and management of data-poor species and fisheries with Australia’s Harvest Strategy Policy. Marine and Coastal Fisheries: Dynamics, Management, and Ecosystem Science 1:244-254

Wilson, J. R., J. D. Prince and H. S. Lenihan. 2010. Setting harvest guidelines for sedentary nearshore species using marine protected areas as a reference. Marine and Coastal Fisheries: Dynamics, Management, and Ecosystem Science. 2:14-27.

Appendix 5 Belcher, C. N. and C. A. Jennings (2010). "Identification and evaluation of shark bycatch in Georgia’s commercial shrimp trawl fishery with implications for management." Fisheries Management and Ecology.

Butler, J. and G. Heinrich (2005). The Effectiveness of Bycatch Reduction Devices o Crab Pots at Reducing Capture and Mortality of Diamondback Terrapins and Enhancing Capture of Blue Crabs. NOAA Project Final Report, University of North Florida: 9.

118 Cox, T., R. L. Lewison, et al. (2007). "Comparing effectiveness of experimental and implemented bycatch reduction measures: the ideal and the real. ." Conservation Biology 21(5): 1155-1164.

Eayers, S. (2007). A Guide to Bycatch Reduction in Tropical Shrimp-Trawl Fisheries, Revised Edition. Rome, FAO 124.

FAO (2009). Guidelines to Reduce Se Turtle Mortality in Fishing Operations. Rome, FAO: 139.

Finkelstein, M., V. Bakker, et al. (2008). "Evaluating the potential effectiveness of compensatory mitigation strategies for marine bycatch. PLoS ONE 3(6): e2480.

Gillet, R. (2008). Global Study of Shrimp Fisheries. Rome, FAO.

Gilman, E. and C. Lundin (2008). Minimizing Bycatch of Sensitive Species Groups in Marine Capture Fisheries: Lessons from Tuna Fisheries, IUCN.

Hall, M. 1998. An ecological view of the tuna-dolphin problem: Impacts and tradeoffs. Reviews of fish biology and fisheries (8):1-34

Løkkeborg, S. (2008). Review and assessment of mitigation measures to reduce incidental catch of seabirds in longline, trawl and gillnet fisheries. Rome, FAO: 33.

SBWG (2010). Report of the Third Meeting of the Seabird Bycatch Working Group. Mar del Plata, Argentina, FAO ACAP.

Žydelis, R., B. P. Wallace, et al. (2009). "Conservation of Marine Megafauna through Minimization of Fisheries Bycatch." Conservation Biology 23(3): 608-616.

Appendix 6 Chuenpagdee, R., L. E. Morgan, et al. (2003). "Shifting gears: Assessing collateral impacts of fishing methods in US waters." Frontiers in Ecology and Environment 1(10): 517-524.

Fuller, S.D., C Picco, et al. 2008. How we fish matters: Addressing the ecologicalimpacts of Canadian fishing gear. Ecology Action Centre, Living Oceans Society, and Marine Conservation Biology Institute. 25pp.

Gillet, R. (2008). Global Study of Shrimp Fisheries. Rome, FAO.

Hall-Spencer, J.M., and P.G. Moore, 2000. Scallop dredging has profound, long-term impacts on maerl habitats. ICES Journal of Marine Science 57: (5) 1407-1415

Icelandic Ministry of Fisheries and Agriculture Fishing Gear-Danish Seine. Accessed on January 20, 2011. http://www.fisheries.is/fisheries/fishing-gear/danish-seine/

Jones, J.B. 1992. Environmental impact of trawling on the seabed: a review. New Zealand Journal of Marine and Freshwater Research. 26: 59-67

119 Kaiser, M.J., J. S.Collie, S. J Hall, S. Jennings, I. R. Poiner. 2001. Impacts of fishing gear on marine benthic habitats. Reykjavik Conference o Responsible Fisheries in the Marine Ecosystem. Reykjavik, Iceland. ftp://ftp.fao.org/fi/document/reykjavik/pdf/12kaiser.PDF

Neckles, H., F.T. Short, S. Barker , B.S. Kopp 2005. Disturbance of eelgrass Zostera marina by commercial mussel Mytilus edulis harvesting in Maine: dragging impacts and habitat recovery. Marine Ecology Progress Series. 285: 57–73

NEFMC. 2010. Essential fish habitat (EFH) omnibus amendment. The swept area seabed impact (SASI) model: A tool for analyzing the effects of fishing on essential fish habitat. Part 1: literature review and vulnerability assessment. Newburyport, MA. 160 pp. http://www.nefmc.org/habitat/index.html

PFMC. 2005. Pacific coast groundfish fishery management plan for the Califonia, Oregon and Washington groundfish fishery. Appendix part 2. The effects of fishing o habitat: west coast perspective. PFMC Portland, OR. 48pp.

Rose, C., A. Carr, D. Ferro, R. Fonteyne, P. MacMullen. 2000. Using gear technology to understand and reduce unintended effects of fishing on the seabed and associated communities: background and potential directions. ICES Working Group o Fishing Haarlem, The Netherlands. 19pp.

Sainsbury, K.J., Campbell, R.A. Lindholm, R., Whitlaw, A.W. (1998 Experimental management of an Australian multispecies fishery: examining the possibility of trawl-induced habitat modification. Global Trends: Fisheries Management (eds E. K.Pikitch, D. D.Huppert & M. P. Sissenwine), pp. 107 112. American Fisheries Society, Bethesda, Maryland

Thrush, S.F., J.E. Hewitt, et al. 1998. Disturbance of the marine benthic habitat by commercial fishing: impacts at the scale of the fishery. Ecological Applications. 8(3): 866–879

Valdemarson, J.W., and P Suuronen. 2001. Modifying fishing gear to achieve ecosystem objectives. Reykjavik Conference on Responsible Fisheries in the Marine Ecosystem. 20pp.

Appendix 7 Auster, P. 2001. Defining Thresholds for Precautionary Habitat Management Actions in a Fisheries Context North American Journal of Fisheries Management 21:1–9

He, P. (2007). Technical measures to Reduce Seabed Impact of Mobile Fishing Gears. Bycatch Reduction in the World’s Fisheries. S. Kennelly: 141-179.

Fujioka, J.T. 2006. A model for evaluating fishing impacts on habitat and comparing fishing closure strategies. Canadian Journal of Fisheries and Aquatic Sciences 63:2330-2342.

Valdemarsen, J., T. Jorgensen, et al. (2007). Options to mitigate bottom habitat impact of dragged gears. FAO Fisheries Technical Paper. 506.

120