A number of scientists from the AFSC participated in the 30th Lowell Wakefield Symposium, held 12-15 May 2015 in Anchorage, Alaska. The Wakefield Symposium, named for an early leader in Alaska’s fishing industry, is organized by the Alaska Sea Grant College Program. The annual symposia cover a different fisheries science topic each year. For 2015, the theme was “Tools and strategies for assessment and management of data-limited fish stocks.” This topic has gained increasing importance as fishery managers in the United States and across the globe strive to ensure sustainable fish populations. Even the most sophisticated stock assessments have some degree of data limitation; for the majority of the world’s fisheries, the lack of data regarding key parameters and quantities poses a major challenge to assessment. Fortunately the global fishery science community has been very active over the last decade in developing statistical methods, many of them quite simple, for determining management quantities (e.g. sustainable fishing mortality rates) when data are scarce. Many of those techniques were on display at the Wakefield Symposium.
Scientists from the REFM division made some important contributions to the meeting’s success:
Dr. Steve Barbeaux discussed progress that he together with colleagues from Korea have been making in using commercial fishing vessels as research platforms for gathering additional data for use in stock assessment. This work is directed towards ameliorating one of the main hurdles to data collection: the cost of field research, particularly chartering research vessels. Steve has been working in Alaska and in Korea to make connections with the fishing industry and explore the types of data collection that can be integrated with the vessels’ existing operations. Successful operations so far include the deployment of oceanographic sensors on fishing gear (e.g. placing temperature and salinity dataloggers on trawl nets to gather information on fish habitats), and recording data from ships’ echosounders (which can be used to enhance population abundance estimates and our understanding of fish distributions).
Dr. Paul Spencer presented the results of a multiyear project to develop a better way of producing biomass estimates for Alaska fish stocks. This is especially important for stocks that are managed under a “Tier 5” approach, where catch limits are based on the estimated biomass multiplied by the natural mortality rate. In the past, this has typically been done by averaging the most recent surveys. Averaging can be problematic in at least two ways: it may cover up trends in the data, and it usually doesn’t account for uncertainty in the estimates. Paul and a team of AFSC scientists explored several different statistical approaches to the problem, and developed a random-walk model that is simple and flexible yet also resolves the problems caused by averaging. The technique is now mandated for use in all of the Tier 5 stocks assessed by the AFSC.
Liz Conners reviewed work she has done on developing a novel technique for setting catch limits for octopus. Octopus are a management concern because biomass estimates are highly uncertain and natural mortality rates are unknown. Relying on the theoretical premise that the natural mortality rate is an acceptable proxy for a sustainable fishing mortality rate, Liz and Kerim Aydin used the predation by Pacific cod on octopus in the Bering Sea to develop octopus catch limits. Using diet and consumption data, as well as the abundance of Pacific cod, they developed an estimate of total annual consumption of octopus in the eastern Bering Sea. They then used this as a proxy for the overfishing level for the octopus population. Their approach is currently being used for setting octopus catch limits in the Bering Sea.
Dr. Olav Ormseth was also involved in the conference. He served on the steering committee and moderated a panel on the management of data-limited fish stocks. Panelists came from Australia, South Africa, and the United States and discussed how data-limited fisheries are dealt with across the world. They were also asked what the “take-home messages” were for the symposium participants. In addition to a general sense that the panelists were impressed with the amount of work being done in this area, there was clear consensus on two points. First, it is necessary to focus on creating and using simple, straightforward methods for dealing with data scarcity (i.e. “keep it simple, stupid”, which was the title of one of the keynote talks and a recurring theme throughout the conference). This reflected the fact that complex models are unnecessary in most data-limited situations, and may actually be inappropriate if there is not enough data to support the model. A second point was the need to consider the perspective of stakeholders (for example, fishers and coastal communities) when developing data-limited approaches. Such approaches need to be transparent and readily understandable, particularly when stakeholders have limited experience with science and management.
