Quality warning
Despite careful data control carried out by Ifremer, inconcistencies inherent to the aggregation of the data might occur. For example :
- It is not recommended to simply combine data from different surveys in an analysis despite the fact that all data from all surveys are presented in the same format. For example, the catchability of the same species can vary with the survey gear used. As a consequence each survey provides a particular but coherent picture of the sampled ecosystem.
- A common feature of survey time series is the change of taxonomic expertise over time. As a result new species can appear, disappear or similar species be confused thus biasing time trends of certain species.
- For any given survey time series, any changes in the sampling procedure, gear, period and area covered might influence the catches. Therefore to avoid the risk of any bias in the analysis, the data must be carefully filtered before use.
It is strongly recommended to use the data with care. In the case of doubts about the validity of the data, please contact the administrator of the data base (harmonie@ifremer.fr).
