Sampling occurred between 5 and 10 meters depth. The sample size was 3 cubic meters. For complete methodology, refer to Richardson et al (2006).
\n
Note: Sometimes, taxa are spotted but cannot be counted (e.g., just a fragment of a specimen was found, presence of mucilage, etc.). In this case, the abundance value was\u00a0set to 1e-10 (provided in \"Data Files\" as \"0.00000000010\" or \"1.0E-10\" values depending upon the data access option selected).
\n
Note: CPR data contains real zeros (we sampled, looked for a specific taxon but did not find it).
\n
Note: The DRI \"DRI or Date of Routine Identification,\" is the year before which a given taxon was not a target for the CPR taxonomists' team. Before this date, abundance entries are provided as missing data values\u00a0(will show as blanks in .csv file).\u00a0\u00a0
\n** At BCO-DMO, Missing data values are displayed differently based on the file format you download.\u00a0 They are blank in csv files, \"NaN\" in MatLab format, etc.
Over the last 90 years, the CPR Survey analysis team has analyzed more than a quarter of a million samples from over 7 million miles of tows in the North Sea, Norwegian Sea, North and South Atlantic, North Pacific, and Indian Oceans.
\n
Spatial and temporal data are stored at the Marine Biological Association of the UK (MBA) for every sample analyzed by the CPR Survey, since 1946. This amounts to over 261,000 samples, with around 200 more samples added per month. The presence of every planktonic entity identified on each sample is stored in the database, and there are over 2 million plankton records in total. The database also contains supportive information such as tow locations, times and dates, ship details, a taxon catalog and analyst details.
\n
Over 800\u00a0zooplankton and phytoplankton entities have been identified on CPR samples, and the 'abundance' of each entity on each sample can be extracted from the database. Some plankton are identified to species level, some to genus level, and some at a higher taxonomic level. Some entities are groups of other entities. The complete Species List is kept in the database.
\n
Data can be extracted from user-defined areas, over specified periods, for selected entities [from the 'The CPR Survey' site]. For example, all samples taken from the Dogger Bank area in the North Sea during March, April, and May since 1946 could be extracted from the database, and the 'abundance' of selected diatom species on each sample could be listed. Alternatively, an average value, number of samples, and standard deviation per year per month could be retrieved. The data can be exported to statistical and presentation packages in many popular formats such as text, rich text, comma-separated, MS Excel, MS Access, MS Word, Fox Pro, Dbase, Lotus, and to SQL compliant databases. The CPR Survey can supply some descriptive data at little cost (usually free).
\n
Future updates are planned to extend the time range of this dataset at BCO-DMO. If you would like to know more about CPR coverage of a particular location, contact David Johns at The CPR Survey.
\n
For information about methods and parameters, and link to The CPR Survey data page: https://www.cprsurvey.org/data/our-data/.
Version differences for the primary data table\u00a0765141_v6_cpr-plankton-abundance.csv (compared to previous version 5 (2024)):
\n
This data update (v6, 2025)\u00a0data has two major improvements over the v5 (2024)\u00a0data:
\n1 - an extra year of finalised sample taxonomy, making the new temporal extent 1958-01-01 to 2022-12-31.
\n2 - an expanded bounding box, from (N:65 to S:35 and\u00a0W:-75 to E:-25) to new bounding box\u00a0(N:65 to S:35 and \u00a0W:-76 to\u00a0E:-22 E).\u00a0 \u00a0All CPR samples available within this bounding box were included. This represents a practical improvement on historical data provision, as the slightly narrower bounding box used in earlier data updates to NSF (above) could 'cut off' the ends of some CPR tows.\u00a0 \u00a0Thus, additional CPR samples from all years since 1958 have been included this time.
\n
Notes on differences for the taxon list between dataset version 5 (2024-08-29) and 6 (2025-07-11, current version):
\n1 - The file \u201c765141_v6_cpr-list-taxa.csv\u201d has a minor difference when compared to the one done last year (\u201cCPRData_UScontract_ListTaxa_02052024.csv\u201d). This is due to changes in the new Matlab database.
\nIn CPRData_UScontract_ListTaxa_02052024.csv , \u201cname_worms\u201d corresponded to the name as extracted from WoRMS. This field does not exist anymore.\u00a0In CPRData_UScontract_ListTaxa_26062025.csv, \u201cWoRMS_Name\u201d corresponds to the name used to extract the Aphia_ID.
\n2 \u2013 The taxa list in 2025 is shorter than in 2024 (413 instead of 414).
\nTaxa 1586 is no longer used in the CPR database.
\nAlso, taxa 399 was replaced by 398.
\n3 \u2013 we generally have a one-to-one match at sample level between extract from 2024 (v5) and the new one.
\nHowever, this is not true for taxa 10703 as it has been modified.