4. Precautionary Approach to Fishery Research

51. Application of the precautionary approach to fishery management depends on the amount, type and reliability of information about the fishery and how this information is used to achieve management objectives. The precautionary approach to fishery management is applicable even with very limited information. Research to increase information about a fishery usually increases potential benefits while reducing the risk to the resource. The scientific and research input that is required for the precautionary approach to fisheries is considered under the following headings; management objectives, observations and information base, stock assessment and analysis and decision processes.

4.1. The Role of Research in Establishing Management Objectives

52. There is a valid scientific role in helping managers develop objectives, so that scientific input to the overall management process is as effective as possible in achieving management intent. The precautionary approach requires continuing and anticipatory evaluation of the consequences of management actions with respect to management objectives. Scientific evaluation of consequences with respect to management objectives requires explicit definition of quantifiable criteria for judgement. An important scientific contribution is in the development of operational targets, constraints and criteria that are both scientifically usable and have management relevance.

53. Research is required to help formulate biological objectives, targets and constraints regarding the protection of habitat, the avoidance of fishing that significantly reduces population reproductive capacity, and reduces the effects of fishing on other (e.g., non-target) species. Combined with biological research, research on socio-economics and the structure of fishing communities is needed to formulate management objectives.

54. Until stock specific research leads to the establishment of alternative operational target based on research and practical experiences, a precautionary approach would seek to: (a) maintain the spawning biomass at a prudent level (i.e., above 50% of its unexploited level), (b) keep the fishing mortality rate relatively low (i.e., below the natural mortality rate), (c) avoid intensive fishing on immature fish, (d) protect the habitat.

4.2. Observation Processes and Information Base

55. A precautionary approach to fisheries requires explicit specification of the information needed to achieve the management objectives, taking account of the management structure, as well as of the processes required to ensure that these needs are met. Periodic evaluation and revision of the data collection system is necessary.

56. A precautionary approach would include mechanisms to ensure that, at a minimum, discarded catch, retained catch and fishing effort data are accurate and complete. These mechanisms could include use of observers and identification of incentives for industry co-operation.

57. Recognizing that resource users have substantial knowledge of fisheries, a precautionary approach makes use of their experience in developing an understanding of the fishery and its impacts.

58. The precautionary approach is made more effective by development of an understanding of the sources of uncertainty in the data sampling processes, and collection of sufficient information to quantify this uncertainty. If such information is available it can be explicitly used in the management procedure to estimate the uncertainty affecting decisions and the resulting risk. If such information is not available, a precautionary approach to fishery management would implicitly account for the unknown uncertainty by being more conservative.

59. Precautionary fishery monitoring is part of precautionary research. It includes collection of information to address issues and questions that are not only of immediate concern but which may reasonably be expected to be important for future generations on in case objectives are changed. Information should be collected on target species, bycatch, harvesting capacity, behaviour of the fishery sector, social and economic aspects of the fishery, and ecosystem structure and function. Measures of resource status independent of fishery data are also highly desirable.

60. The precautionary approach relies on the use of a history of experience with the effects of fishing, in the fishery under consideration and/or similar fisheries, from which possible consequences of fishing can be identified and used to guide future precautionary management. This requires that both data and data collection methods are well documented and available.

61. There are many management processes and decision structures used throughout the world, such as regional management bodies, co-management, community-based management, and traditional management practices. Research is needed to determine the extent to which different management processes and decision structures promote precaution.

4.3. Assessment Methods and Analysis

62. Biological reference points for overfishing should be included as part of a precautionary approach.

63. A precautionary approach specifically requires a more comprehensive treatment of uncertainty than is the current norm in fishery assessment. This requires recognition of gaps in knowledge, and the explicit identification of the range of interpretations that is reasonable given the present information.

64. The use of complementary sources of fishery information should be facilitated by active compilation and scientific analysis of the relevant traditional knowledge. This should be accompanied by the development of methods by which this information can be used to develop management advice.

65. Specifically the assessment process should include:

a. scientific standards of evidence (objective, verifiable and potentially replicable), should be applied in the evaluation of information used in analysis;

b. a process for assessment and analysis that is transparent, and

c. periodic, independent, objective and in-depth peer review as a quality assurance.

67. Sources of uncertainty in data include: (a) estimates of abundance; (b) model structure; (c) parameter values used in models; (d) future environmental conditions; (e) effectiveness of implementation of management measures; (f) future economic and social conditions; (g) future management objectives, and (h) fleet capacity and behaviour.

68. Specific alternative hypotheses about underlying biological, economic and social processes to be considered include: (a) depensatory recruitment or other dynamics giving rapid collapse; (b) changes in behaviour of the fishing industry under regulation, including changes in coastal community structure; (c) medium-term changes in environmental conditions; (d) systematic underreporting of catch data; (e) fishery-dependent estimates of abundance not being proportional to abundance; (f) changes in price or cost to the fishing industry; and (g) changes in ecosystems caused by fishing.

69. In calculating (simulating) the response of the system to a range of alternative management actions, the following should be taken into account:

a. short-term (1-2y) projections alone are not sufficient for precautionary assessment; time frames and discount rates appropriate to inter-generational issues should be used, and

b. scientific evaluation of management options requires specification of operational targets, constraints and decision rules. If these are not adequately specified by managers, then precautionary analysis requires that assumptions be made about these specifications, and that the additional uncertainty resulting from these assumptions be calculated. Managers should be advised that additional specification of targets, constraints and decision rules are needed to reduce this uncertainty.

a. where there are no sufficient observations to assign probabilities to different states of nature that have occurred, decision tables could be used to represent different degrees of management caution through the Maximin and Minimax criteria;

b. where the number of different states of nature and the number of potential management actions considered are small, but probabilities can be assigned, decision tables can be used to show the consequences and probabilities of all combinations of these, and

c. where the range of states of nature is large, the evaluation of management procedures is more complex, requiring integration across the various sources of uncertainty.

72. Since concern regarding the reversibility of the adverse impacts of fishing is a major reason for a precautionary approach, research on reversibility in ecosystems should be an important part of developing precautionary approaches.

4.4. Implementation Guidelines

73. The following measures could be applied in order to implement a precautionary approach to fishery research:

a. take into account the best scientific evidence available when designing and adopting management and conservation measures, in accordance with the provisions of the 1982 UNCLOS Convention. In the context of precautionary management, the best scientific evidence is described in section 4.3;

b. require a minimum level of information to be made available for any fishery to start or continue;

c. ensure that the “lack of full scientific certainty shall be not used as a reason for postponing cost-effective measures to prevent environmental degradation” (principle 15 of the Rio Declaration);

d. reduce critical uncertainties in the management plan;

e. take measures aimed at eliminating or reducing non-reporting and misreporting, inter alia, by ensuring that the fisheries sector cooperates in data collection and the public is fully informed of the results and uncertainty in the assessment;

f. systematically analyze various possible management options using the whole range of available models (bioeconomic, multispecies and behavioural), showing: (a) the likely direction and magnitude of the biological, social and economic consequences; (b) the related levels of uncertainty and the potential costs of the proposed action (risk assessment), and of no action (status quo scenarios);

g. promote multidisciplinary research, including: (a) social, economic and environmental sciences, and (b) research on management institutions and decision-making processes;

h. develop scientific information on multispecies and ecosystem processes as a foundation for identifying acceptable degrees of disturbance;

i. identify biological limit and target reference points for affected species and stocks, habitats and the ecosystem at large;

j. identify bioeconomic reference points to address the objectives of the fishery management plan;

k. improve methods for quantification of direct and indirect impacts of fishing;

l. improve understanding of the performance of different management structures in relation to precaution;

m. develop methods for optimizing the monitoring system, and

n. develop research and development programmes aiming at improving the performance of fishery technology in relation to environmental impacts and precautionary management.

74. The Minimax/Maximin approach is a way of examining uncertainty and guiding decisions without the need for explicit statements of probabilities on the alternative hypotheses (Schmid, A. 1989, Cost-benefit analysis, West View Press). In the table below, S1 and S2 represent the alternative hypotheses about the resource (sometimes called “the different states of nature”). In this example, S1 is a hypothesis that implies a higher level of resource productivity and sustainable yield than S2. In the table rows, D1 to D3 represent alternative decisions. In this example D1, D2 and D3 broadly represent respectively a high, medium, and low level of fishing effort. The Ps represent the probabilities being placed on hypotheses being true. Values in the table represent the relative value of the outcome of a decision as applied to a given state of nature. In the example, these values could be regarded as representing sustainable catch.

Decision	S1 p=?	S2 p=?
D1	100	5
D2	50	40
D3	70	20

75. The Maximin Values Criterion is a cautious approach that leads to selects the maximum (highest) of the minimum outcome. The following table gives the relative value of outcomes for decisions given the hypothesis being true. Decision 2 would be chosen by this approach.

Decision	S1 p=?	S2 p=?	Minimum Value
D1	100	5	5
D2	50	40	40
D3	70	20	20

76. The Minimax Regret Criterion is a less cautious approach that selects the minimum of the maximum regret. The following table give a measure of regret for each decision when the hypothesis is true. Decision 3 would be chosen by this approach.

Decision	S1 p=?	S2 p=?	Minimum Regret
D1	100-100=0	40-5=35	35
D2	100-50=50	40-40=0	50
D3	100-70=30	40-20=20	30

77. The decision table approach uses the probabilities of alternative hypotheses of the state of nature, along with the values of the outcomes of decisions to give an expected value and variance of each action across all alternative hypotheses.

Decision	S1 p=0.7	S2 p=0.3	Expected Value	Variance
D1	100	5	71.5	1895.25
D2	50	40	47.0	21
D3	70	20	55.0	525

78. The decision is to select the desired trade-off between the variance and the expected value of the outcome. The variance is a measure of uncertainty and the expected value gives the expected result of choosing a given policy. Policy D1 has a high expected outcome with the highest uncertainty, and a 0.3 probability of a very poor outcome. On the other hand, policy D2 has a much lower expected outcome with lower uncertainty, and no probability of a poor outcome.

79. Methods that integrate across uncertainty: Where the number of possible states of nature is large, as is almost always the case, a mathematical equivalent operation to that described above for a simple decision table can be carried out. This results in the calculation of outcome probabilities for each possible decision, integrating across uncertainties.