This project was supported by a portion of the $54,597 in IPHC research funds (FY 1994/1995) for computer programming. Project reference number for FY 1995/1996 is 1004.
A standardized SQL database is established containing all observer bycatch data from the domestic groundfish since the inception of the Domestic Observer Program in 1990. The database will provide a common reference for all bycatch researchers, and can be updated annually.
As halibut bycatch in the domestic groundfish fisheries persists as a significant source of halibut mortality, increasing demands are being placed on the data. In 1995, IPHC staff used data collected on-board groundfish vessels for several studies related to bycatch management and analysis. Beginning in 1996, bycatch will be incorporated into the annual stock assessment. Data are transferred from the Observer Program around August each year for the previous year's sampling. These data are passed through a series of FORTRAN programs to calculate various quantities of interest and are reformatted for use in statistical computer programs. The format of the observer data has changed every year, thus the FORTRAN programs are modified each year. None of the annual datasets are perfectly aligned with the others, which greatly complicates historical analyses of observer data. Additionally, the annual datasets that are prepared are static, i.e., once prepared they are not updated. Revision of the observer program data is an ongoing process, thus data obtained in 1990 may have undergone significant alteration.
Because of the growing importance of the observer bycatch data, it was deemed crucial to develop a high-quality database. The primary objectives were to create an historical database that
1) is standardized across years
2) is expandable
3) is accessible to all IPHC staff working on bycatch
4) is in a format amenable to inclusion in the annual stock assessment
The entire observer database relating to bycatch in the domestic groundfish fisheries dating to the beginning of the domestic program in 1990 was requested and obtained by August. The data were transformed to two tables within a Microsoft Access SQL database. The haul table contains comprehensive retrieval and species composition data on 217,000 individually sampled hauls (Table 1). The incidence of halibut is recorded with a large array of physical and biological variables that can be analyzed for systematic bycatch patterns. The length table contains length and/or viability data on 760,000 incidentally captured halibut (Table 2). The data underwent comprehensive screening for errors. Particular attention was paid to the capture locations. All hauls were plotted with a GIS package; this simple step resulted in the location correction of several thousand hauls that had been miskeypunched. Many other basic errors were also identified and corrected concerning fishing depth, gear type, fishing times, numbers of halibut in a haul, vessel codes, etc. This fine attention to detail was possible due to the SQL query ability of Access, hinting at the powerful type of data querying now possible with this database.
Following correction of data entry errors, the data are then processed via internal programs. Computed quantities of interest include bycatch viability by target fishery, and size distribution of halibut mortality by region and target fishery. These programs will allow for semi-automatic calculation of bycatch size distributions required for the annual stock assessment beginning in 1997.
Table 1. Number of individual hauls sampled for species composition and halibut bycatch in the domestic groundfish fisheries.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Table 2. Number of individual bycaught halibut sampled for length and viability in the domestic groundfish fisheries.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|