Aside the main db report, which describes all the vocabularies used in this document, we have to build a separate database for metrics. In the project these correspond to LHT, but it’s more, the data should also correspond to the time series and the sampling db developped by WGEEL. The two latter data structure (series and sampling) are very similar and they both hold very similar group metrics and individual metrics.
The first was developed initially to store data about the series used in recruitment. In practice, it consists of three tables, the t_series_ser (Figure Figure 1 - top in blue) table contains series id and description, with columns describing the sampling details, the stage used, the method… This is the main identifier of the series which will be used as a reference in all dependent tables. The second t_dataseries_das table (Figure Figure 1 - on the right) holds data about annual values in series. These are typically annual counts for recruitment, along with additional effort data. Linked to these are group metric series used to describe the series, mean age of eel, mean size, proportion of glass eel among the yellow eels, proportion of females … (Figure Figure 1 - in orange) Finally, we can link individual metrics. The individual metrics are all detailed for one fish. And they concern metrics like size, weight, sex, but also can hold data about quality, contamination. So these are in essence the Life History traits analysed by WP2 in DIASPARA (Figure Figure 1 - in pink).
The second type of data was developed to hold the data collected for DCF. These can be metrics collected from sampling by the fishermen, data coming from the analysis of electrofishing data, or other experimental sampling that are not reported as series. Currently the two structures for series and sampling are very close, the only difference is that there is no annual number linked to the sampling data, and that they are not linked to a stage in the first table, so the stage is added in the fish table. The difference in table structure is illustrated below in tables highlighted in yellow (Figure Figure 2).
The database development highlighted in the current report has several objectives :
The first objective is to join the two database to simplify the database development and handling of data.
The second objective is to use the new referentials created for the migdb database.
The third objective is to import data from WP2, the excel sheets have been created in february 2025 and will already (in March) require some adaptation as the database evolves, for instance the referential of stages is no longer in line with the templates.
The fourth objective is to adapt the ICES eel format creating during the first quarter of 2025 for eel, to integrate data with DATSU. During discussions with ICES, it was decided to postpone the integration of wgeel database in DATSU as this will require an adaptation of the shiny scripts of data integration.
The fifth objective is to hand over this database, along with the migdb to ICES, for integration in ICES database ecosystem.
Creating the database structure from WGEEL (TODO)
git issue #23 Write simplified structure from WGEEL
The main issue will require to merge the two table structures (sampling and series) and adapt to migdb vocabulary.
Once done a beta version probably not completely adapted will be released.
Import data from WGEEL (TODO)
If this works then the rest should follow smoothly git issue #25 Import data from WGEEL
The release date for that one is :
So the metric release will be after wgeel, but hopefully some of the work will be started and this can be discussed during wgeel.
