North Sea Benthos Project 2000 Data Management
Vanden Berghe, E.; Rees, H.L.; Eggleton, J.D. (2007). North Sea Benthos Project 2000 Data Management. CM Documents - ICES, CM 2007(A:18). ICES: Copenhagen. 15 pp. Part of: ICES CM Documents - ICES. ICES: Copenhagen. ISSN 1015-4744, more |
| Authors | | Top | - Vanden Berghe, E., more
- Rees, H.L., more
- Eggleton, J.D.
| | |
| Abstract | Data for the ICES North Sea Benthos Project 2000 were provided by 15 organisations. Also the data from the 1986 North Sea Benthos Survey were available for analysis and comparison with the 2000 data sets. Sampling occurred mainly in spring and early summer of 2000 and covered almost the whole North Sea from the English Channel to about 60°N. All biogeographic data were integrated in a single database, resulting in a dataset with 91,362 distribution records, from 1570 stations, 2609 samples, and 1954 different taxa. Great care was taken to standardise taxonomic names used, and to deal with incomplete and uncertain identifications in a clear and consistent way. Preliminary plotting and analysis made it clear that there are large biases in sampling locations, resulting from the needs and objectives of the original data collector. Sample size was not always recorded, and had to be reconstructed from the data in some cases. Also, sample size was very uneven, and necessitated extra caution in interpretation in the results of the analysis. However, a comparison of 1986 data (where there was an agreed-upon plan before sample collection) and 2000 data (where data were used opportunistically) showed that patterns were consistent between the two sampling periods, and hence that the opportunistic nature of the 2000 dataset did not invalidate any possible science. Obviously, integrated data sets like the one presented will never be a substitute for carefully planned joint research, but it is argued that there is great merit in exercises such as this. Existing data are given a second lease of life, thereby greatly increasing their value, without going to the cost of collecting new samples. Integration of data offers an opportunity for an extra quality control, and can bring to light potential problems not detected at first analysis. Last but not least, integration will lead to much larger data sets than what can be achieved with centrally-planned research; it leads to datasets commensurate with the scale of environmental problems humankind is presently confronted with. |
|
