Considerations needed for analysing data from the Swedish Electrofishing RegiSter (SERS), with special reference to the RivFishTIME database of long-term riverine fish surveys

The published database RivFishTIME (Comte et al. 2021, Global Ecology and Biogeography, doi: 10.1111/geb.13210) includes a large section of time-series data on fish abundance in Swedish rivers from the Swedish Electrofishing RegiSter, SERS. Knowledge about the limitations of the source data are important when extracting and analyzing data and with this brief note we provide some details that may be helpful for interpreting the Swedish time-series. The note highlights the importance of linking vital metadata to extracted focal data when constructing new databases, especially concerning time series data from monitoring programs conducted in non-randomly selected sites with human environmental impacts. Many of the SERS data come from rivers that have been affected by human impact, e.g. liming to mitigate environmental acidification and hydropower dams, since before monitoring was initiated. Data in SERS are also biased towards shallow salmonid habitats, due to the configuration of Swedish monitoring programs. Hence, data from many rivers are not representative of their fish biodiversity in general. This information is vital for appropriate interpretation of fish biodiversity trends. For RivFishTIME analyses considerations are important since Swedish data constitutes a large proportion of the database. We also provide background information about SERS and references to other Swedish databases containing complementary information. Finally, we provide contact information of the SERS database  curators, who can assist prospective analysts with data extraction from SERS.