Grant: Discovery Grants Program - Individual

The fishery is worth $6B to the Canadian economy, $1.6B in Atlantic Canada, and employs thousands. One in five Canadians lives in a coastal community and ocean industries generate over $30B a year. Thus the health of the oceans' resources is critical. However, in recent years major Northwest Atlantic fish stocks such as Atlantic cod and American plaice crashed, and there has been no clear sign of recovery since. This is an economic and environmental concern for Canadians. Data plays a crucial role in understanding the current state of ocean ecosystems and predicting their future. For more than 40 years, Fisheries and Oceans Canada (DFO) amassed huge data stores from sampling of fisheries landings, and annual/seasonal multispecies surveys. Other countries also conduct surveys. These datasets are complex due to intricate sampling strategies that change over time, incomplete survey coverage due to poor weather and mechanical problems with vessels, under-reporting of catches, and complicated fish activities like seasonal migration and schooling. Statistical methods that do not properly account for these complexities can lead to misinterpretation of data, misleading predictions and, ultimately, suboptimal fisheries management strategies. The goals of the proposed program are to 1) create new and better statistical tools for mining complex datasets and 2) use these tools to extract the information needed to understand, predict, manage, and protect Canadian fisheries and ecosystems. Our novel statistical models/methods and quantitative techniques will better address complicated sampling, various data types, complex species-specific behaviors, and environmental factors. These improved methods will also account for important life-history traits such as the effect of maturity-induced energy intake relocation on fish growth. Where data span decades and large spatial scales, we will employ a spatiotemporal modeling framework and develop efficient modeling methods and algorithms. Data is crucial to develop and test statistical methods. This research program, enabled by complex fisheries data, will enrich statistical theory and application. The models/methods we develop will have wide application in fisheries and ecology, and can be extended to areas with similar data types and sampling designs such as economics, medicine, etc. Through this program, HQP will develop advanced quantitative skills for spatiotemporal modeling of complex data. They will collaborate with national/international experts in statistical modeling of fisheries and ecological data, and will develop open access codes and packages that will facilitate the uptake and impact of their research achievements. HQP will be prepared to pursue advanced research in statistics and to assist governments and the fishing industry to maintain a sustainable and prosperous fishery. These skills are also in demand in other ecological and industrial sectors internationally and in Canada.