Journal cover Journal topic
Proceedings of the International Association of Hydrological Sciences An open-access publication for refereed proceedings in hydrology
Journal topic

Journal metrics

Journal metrics

  • CiteScore value: indexed CiteScore
    indexed
  • SNIP value: indexed SNIP
    indexed
  • SJR value: indexed SJR
    indexed
  • IPP value: indexed IPP
    indexed
Volume 379
Proc. IAHS, 379, 169-174, 2018
https://doi.org/10.5194/piahs-379-169-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.
Proc. IAHS, 379, 169-174, 2018
https://doi.org/10.5194/piahs-379-169-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Pre-conference publication 05 Jun 2018

Pre-conference publication | 05 Jun 2018

Detecting trend on ecological river status – how to deal with short incomplete bioindicator time series? Methodological and operational issues

Flavie Cernesson1, Marie-George Tournoud2, and Nathalie Lalande3 Flavie Cernesson et al.
  • 1TETIS, AgroParisTech, Cirad, CNRS, IRSTEA, Université de Montpellier, Montpellier, France
  • 2HSM, Université de Montpellier, CNRS, IRD, Montpellier, France
  • 3TETIS, IRSTEA, AgroParisTech, Cirad, CNRS, Université de Montpellier, Montpellier, France

Abstract. Among the various parameters monitored in river monitoring networks, bioindicators provide very informative data. Analysing time variations in bioindicator data is tricky for water managers because the data sets are often short, irregular, and non-normally distributed. It is then a challenging methodological issue for scientists, as it is in Saône basin (30000km2, France) where, between 1998 and 2010, among 812 IBGN (French macroinvertebrate bioindicator) monitoring stations, only 71 time series have got more than 10 data values and were studied here. Combining various analytical tools (three parametric and non-parametric statistical tests plus a graphical analysis), 45 IBGN time series were classified as stationary and 26 as non-stationary (only one of which showing a degradation). Series from sampling stations located within the same hydroecoregion showed similar trends, while river size classes seemed to be non-significant to explain temporal trends. So, from a methodological point of view, combining statistical tests and graphical analysis is a relevant option when striving to improve trend detection. Moreover, it was possible to propose a way to summarise series in order to analyse links between ecological river quality indicators and land use stressors.

Publications Copernicus
Download
Short summary
For water managers, analysing time variations in short and incomplete bioindicator series is necessary but tricky. From our work on benthic invertebrate indicator series sampled in the Saône basin (France), we conclude that (1) combining statistical tests and graphical analyses is relevant for classifying time series, (2) environmental contexts explain similarities between series better than river size classes, (3) a single value can summarise a serie to link bioindicators to land-use stressors.
For water managers, analysing time variations in short and incomplete bioindicator series is...
Citation
Share