The effectiveness of using CPUE data derived from Vessel Monitoring Systems and fisheries logbooks to estimate scallop biomass

Lee G. Murray 0 Hilmar Hinz 0 Natalie Hold 0 Michel J. Kaiser 0 0 School of Ocean Sciences, Bangor University , Askew Street, Menai Bridge, Anglesey, LL59 5AB , UK Obtaining accurate data on abundance is vital to undertaking viable stock assessments of commercially exploited species. Satellite Vessel Monitoring Systems (VMSs) combined with fisheries logbooks have the potential to provide an abundant source of data with greater spatial and temporal coverage than research surveys. However, to date it has not been demonstrated how well VMS-derived abundance or biomass indices reflect research survey results. In this study we compared biomass indices of scallops derived from (i) fishing vessel surveys, (ii) research vessel surveys, and (iii) fishery-dependent data using VMSs and logbooks. In most cases there were strong relationships between biomass indices of Pecten maximus from fishing vessels and the research vessel. There were stronger relationships between P. maximus biomass indices from fishery-dependent VMS and logbook data and research vessel data at the beginning of the fishing season, when abundance was higher, but weaker relationships at the end of the fishing season. The time and location of sampling affected biomass estimates over short periods, and without standardizing to location and vessel, biomass depletion was masked. Fishery-dependent data provides a valid means of assessing relative scallop abundance and may prove equally viable in other fisheries with appropriate standardization of Catch Per Unit Effort (CPUE) data, making real-time management of fisheries increasingly feasible. - Satellite Vessel Monitoring System (VMS) data can be used to define fishing grounds (Gerritsen and Lordan, 2011; Jennings and Lee, 2012), to assess the impacts of fishing activity on benthic habitats (Witt and Godley, 2007; Hinz et al., 2009; Lambert et al., 2011; Lambert et al., 2012), to examine fisher behaviour (Murray et al., 2011), and ultimately to improve fisheries management (Chang, 2011). Since VMS records do not provide a direct measure of fishing effort, some specific considerations need to be addressed when using VMS data. As such, several studies have examined different methods of processing and analysing VMS data to estimate fishing effort (Mills et al., 2007; Walter et al., 2007; Lambert et al., 2012). However, VMS data are not yet widely used in stock assessments, due largely to the lack of time-series of sufficient length (ICES, 2011). An ICES study group was established to address the need for a structured approach to storing and accessing VMS data (ICES, 2010). Furthermore, VMS records can be joined to fisheries logbooks to provide more accurate estimates of fishing effort (Deng et al., 2005; Bastardie et al., 2010; Lee et al., 2010; Gerritsen and Lordan, 2011; Murray et al., 2011; Lambert et al., 2012). Thus a VMS tools package has been created to allow VMS data to be linked to logbook data, and to facilitate processing of these data (Hintzen et al., 2012). As longer time-series of VMS and logbook data become available, fishery-dependent Catch Per Unit Effort (CPUE) data are likely to be used much more regularly to assess the status of populations, as it is a cost-effective method of providing year-round data across the majority of a fishing fleet. Therefore, it is important that the accuracy of these data, at least relative to traditional survey methods, is verified. Obtaining accurate abundance indices is a prerequisite to undertaking useful stock assessments. Although the collection of CPUE data is relatively simple, they may not exhibit a linear relationship with abundance. Hyperstability, where abundance declines more rapidly than CPUE (Hilborn and Walters, 1992), is a common property of CPUE indices (Harley et al., 2001), due largely to targeting behaviour of fishers, technological change or vessel effects (Bishop, 2006; Quirijns et al., 2008; Erisman et al., 2011). Therefore, VMS- and logbook-derived data may need to be corrected for these variables before being used as abundance indices. The standardization of CPUE data can have a significant effect on estimates of relative abundance (Carruthers et al., 2011). Underfitting of standardization models to fishery-dependent data can also lead to biased estimates (Ye and Dennis, 2009), and a lack of data from unfished, or rarely fished, areas over which a population extends can lead to inaccurate population size estimates (Campbell, 2004). There are also problems associated with using fisheryindependent data. Although fishery-independent surveys can be designed to avoid bias and reduce uncertainties, conversion coefficients may be required when different vessels are used (Dare et al., 1994; Pelletier, 1998). Moreover, due to their expense, research vessel surveys usually sample only a relatively small area of a fishery during a limited number of sampling events. Therefore, both fishery-dependent and fishery-independent CPUE estimates may differ from true abundance. Consequently, VMS and logbook data are potentially a valuable source of information on the status of exploited populations due to the quantity of data but require appropriate processing and standardization. The availability of VMS and logbook data provides an opportunity to respond to the results of stock assessments with spatial management at a local and regional level. At a local level, vessels could be directed to the most profitable (high abundance) areas and advised against or prevented from fishing in low-abundance areas before abundance reaches uneconomical levels. The margins above these uneconomical levels could be governed by wider-scale stock assessments. Even if total allowable catches (or equivalent caps) are not strictly adhered to, such an approach should facilitate more efficient exploitation of scallop stocks than at present. The aim of this study was to assess the feasibility of using VMS and logbook data to obtain biomass indices for scallops. We examined three methods of obtaining biomass indices in two separate studies. First, research vessel abundance indices for king scallops, Pecten maximus, were compared with those derived from fishing vessel-based surveys. Second, P. maximus biomass estimates from research vessel surveys were compared with estimates derived from fishery-dependent VMS and logbook data. The effectiveness of standardizing fishery-dependent CPUE data was also examined. Material and methods Fishery-independent surveys Surveys of scallop populations have been undertaken around the Isle of Man since 1992 (see Beukers-Stewart et al., 2003). Up to 12 spatially discrete stations have been surveyed annually in May/June and September/October (Figure 1, abbreviations in brackets): Bradda Offshore (BRO), Bradda Inshore (BRI), Peel (PEL), Targets (TAR), Point of Ayre (POA), Ramsey (RAM), Maughold (MAG), Laxey (LAX), East of Douglas (EDG), South East of Douglas (SED), South of Port St Mary (PSM) and Chic (...truncated)