Background
On the 1st of January 2014, the European Union introduced a phased discard ban or “Landing Obligation” for regulated species, as part of Common Fisheries Policy (CFP) Basic Regulation (Article 15). Today, in the post-Landing Obligation world, it may appear that a report on discard survival has little relevance. However, The Landing Obligation policy includes a high survival exemption (HSE) for “species for which scientific evidence demonstrates high survival rates, taking into account the characteristics of the gear, of the fishing practices and of the ecosystem” (Article 14, paragraph 4b). The HSE generated considerable interest from stakeholders, who wished to demonstrate that their particular fisheries did in fact have a suitably high survival rate for discarded unwanted catch.
Research aimed at determining whether aquatic organisms survive after being caught and subsequently released has been conducted over many decades. However, in 2014, there had been no comprehensive assessment of all the scientific methods and approaches that can be employed to estimate the survival of discarded fish and other aquatic animals. To that end, ICES established the Workshop on Methods for Estimating Discard Survival (WKMEDS), in January 2014, to provide guidance on best practice for methods to quantify the survival of discarded, unwanted catch.
WKMEDS published its first preliminary guidance on best practice for survival estimation methods in April 2014 (ICES, 2014) – just four months after the group was established. This preliminary guidance provided the framework for WKMEDS to develop and apply these methods, over the following years, to gather evidence in support of applications for HSEs. This Cooperative Research Report (CRR) is the culmination of that collective research, providing a state-of-the-art review of the current best practice for methods to estimate discard survival.
Survival Assessment Methods
The guidance presented in this CRR identifies three main approaches for conducting discard survival assessments:
i) Vitality assessments: the vitality of the subject to be discarded is scored relative to any array of indicators (e.g. activity, reflex responses, and injuries) that can be com-bined to produce a vitality score. These assessments do not in themselves generate an absolute survival estimate, but can quantify “at-vessel” or “immediate” mortality lev-els. Where vitality scores have been correlated to the likelihood of survival, by con-ducting vitality assessments in combination with captive observation and/or tagging, they can be used as a proxy for survival likelihood (Section 8).
ii) Captive observation: the subjects undergo a normal catch-and-sorting process, and are then held in captivity for a sufficiently long period to determine their fate (Section 9).
iii) Tagging: the subjects to be discarded are tagged and released, and either their behav-iour/physiological status is remotely monitored (via biotelemetry) to determine its post-release fate, or survival estimates are derived from the number of returned tags (Section 10).

Vitality assessments cannot provide survival estimates in isolation. Captive observation methods can do so, but cannot determine the influence of predation on the survival of discarded organisms. Therefore, the method which can potentially generate the most robust estimates of discard survival is tagging, since it includes the effects of predation.
A wide variety of species may be important as predators of discarded animals, including birds, marine mammals, fish, and even benthic invertebrates. Due to the stresses associated with capture and release, most discarded animals are likely to be compromised in their ability to evade such predators. As such, mortality due to this increased risk of predation may add to any estimates of post-release mortality resulting from the capture and release related stressors alone.
To address this potential source of uncertainty and bias in discard survival estimates, this CRR also describes over the different sections methods for assessing the nature and magnitude of seabird predation on the discarded catch. There are several reasons for focusing on seabird predation including: (i) seabirds forage at or just below the sea surface and, therefore, they are one of the first predatory threats to discarded fish, and can be visually observed with relative ease, (ii) they are arguably the most abundant discard scavengers in many fisheries, and (iii) there is a large body of research investigating seabird–fishery interactions, particularly with respect to the importance of discards as food for avian fauna.
In isolation, each of these methods (vitality assessment, captive observation, tagging observation, and avian predation) has limitations which can restrict the usefulness of the survival estimates they produce. However, when two or more of these methods are combined, there is clear potential for considerable synergistic benefits. The benefits of this integrated approach include: (i) reducing resource requirements, (ii) increasing the scope of the investigation, and (iii) improving the accuracy, precision, and applicability of discard survival estimates.
Selecting the appropriate method/s
The choice of method depends primarily on the objectives for the survival assessment, which should be defined in consultation with key stakeholders in the fishery of interest, including fishers and managers, as well as those conducting and funding the work. This CRR identifies six common objectives ranging from: 1) providing estimates of the proportion of discards that appear dead or impaired at the point of discarding under particular conditions (referred to as “survival potential”); to 6) generating a discard survival rate for a population that is representative of a fishery (management unit), including the influence on survival of selected variables. The methods necessary to achieve these objectives increase in complexity, from the first (1) to the last (6), as do the required amounts resources and time. This will likely influence the initial choice of objective.
Once the objectives have been agreed upon, and the appropriate method(s) selected, the experimental design should ensure that the results will be scientifically valid, while optimizing the use of resources and subject animals. This will be best achieved by considering the following general relevant research questions, in addition method-specific questions and considerations, which are detailed in the relevant sections in the CRR.
• Are clear, unambiguous criteria being used to assess the survival status of test subjects?
• Before investing in a full survival assessment, have test-control subjects been successfully collected, and can they be monitored using the selected method (e.g. captive observation or tagging) without killing a substantial proportion of them?

Will the controls be measurable, predictable, and representative of the experimental population and monitoring methods? • Are the controls representative of the treatment groups; i.e. biologically (length, sex, fitness/condition), and in terms of sample size, and spatial and temporal origin?
• Will the control subjects experience the same experimental conditions?
• Are “blind controls” going to be used to avoid observation bias?
• Have all potential sources of experimental bias been considered in the design and acquisition of the control subjects?
• Have pathway analysis and literature reviews been used to identify the most likely potential explanatory variables to be considered in an experimental design?
• Will treatment subjects and experimental conditions be representative of the fishery?
• Will the treatment and control subjects be randomly selected to account for bias?
• Has an appropriate statistical analysis of the survival data been identified?
• Are appropriate power analysis techniques being used to calculate optimal sample sizes and number of replicates for the hypothesized treatment effects and required statistical power?
Conclusions and Recommendations
The introduction of the Landing Oblogation, and its inclusion of HSE, has motivated stakeholders in several fisheries to begin research programmes, based on the guidance provided in this CRR, to investigate the survival potential of various components of the unwanted catch. This has increased our collective understanding of the stressors associated with the capture process, and, in turn, motivated efforts to mitigate these stressors to promote high discard survival and catch welfare in general.
To further promote development in this field WKMEDS recommends that future survival assessments should:
• be representative of the discarded catch and practices in commercial fishing, ideally at a metier scale;
• utilise one or more of three methods: vitality assessments, captive observation, and/or tagging observation;
• clearly define the state of death for the species of interest, preferably using multiple criteria pertinent to the chosen method;
• avoid biasing results through observation induced mortality, and, wherever possible, demonstrate this using appropriate controls;
• monitor the subject animals at a frequency and over a sufficient time period to describe any delayed mortality attributable to the catch-and-discarding process;
• incorporate careful design and analysis to ensure that the results are scientifically valid, while optimizing the use of resources and subject animals; and
• where practical, integrate two or more of the methods to estimate discard survival, to increase the scope of the investigation, as well as improve the accuracy, precision, and applicability of the estimates.