Date: 2014-01-15 02:00 PM
Speaker: ??? Extended/EventDirector/Description ???
Local: ??? Extended/EventLocation/Company ??? ??? Extended/EventLocation/Street ??? ??? Extended/EventLocation/City ???
Email contact: ??? Extended/EventDirector/Email ???
Phone contact: ??? Extended/EventDirector/Phone ???
Discard estimation is a major topic of the European Fisheries Data Collection Framework. Through time, national sampling programs have evolved from ad hoc sampling designs focused on the precise estimation of single species to probability-based sampling designs focused on accurate estimation of a wide array of species. In Portugal, an onboard sampling programme routinely collects samples from the whole catch of several fisheries but sampling design and sampling effort allocation were originally aimed at the estimation of European hake discards. In our study, we analyze data from the Portuguese bottom otter trawl fishery that targets crustaceans in ICES Division IXa (years 2004-2012) and check it for biases that could hamper the estimation of discards from other species, such as Norway lobster, deepwater rose shrimp and blue whiting. We start by identifying the major spatial and temporal factors that drive catches of the different species in this fishery. We then test a set of raising algorithms in the estimation of landings from this fishery and use them to build time series of discards. We show that spatiotemporal biases in sampling design and sampling effort allocation may severely impact final estimates obtained for some non-target species but that these may be attenuated by a careful consideration of raising algorithms and combinations of estimates at the time of estimation.