Developing an improved sediment-specific biomonitoring tool: Combining expert knowledge and empirical data

Matthew Turley, Gary Bilotta, T. Krueger, R.E. Brazier, Chris Extence

Research output: Contribution to conferenceAbstract


The Proportion of Sediment-sensitive Invertebrates (PSI) index is a biomonitoring tool designed to identify the degree of sedimentation in rivers andstreams. The index has a sound biological basis, using invertebrate sensitivity ratings that were determined following an assessment of faunal traits associated with a sensitivity or tolerance of fine sediment. Despite a moderate correlation with deposited fine sediment, comparable to other pressure-specific indices used for Water Framework Directive classification, the large variability in the relationship limits confidence in its application. In this study, sediment and invertebrate data, collected from a range of reference condition river and stream ecosystems (n=2252), is used to empirically-assign species sensitivity weights in an attempt to improve the performance of the PSI index. To maintain the index's biological basis, sensitivity weights were restricted to a range, based on their original sensitivity ratings. The optimum set of sensitivity weights were identified using non-linear optimisation, as those that resulted in the highest Spearman's rank correlation between Empirically-weighted PSI (E-PSI) scores and deposited fine sediment. Applying these optimum sensitivity weights to an independent test dataset (n=252) showed E-PSI to have a strongcorrelation with deposited fine sediment (rs=-0.74,p<0.01), compared to a moderate correlation for PSI (rs=-0.66,p<0.01).
Original languageEnglish
Number of pages1
Publication statusPublished - 8 Jul 2015
Event9th Symposium for European Freshwater Sciences - University of Geneva, 5-10 July 2015
Duration: 5 Jul 2015 → …


Conference9th Symposium for European Freshwater Sciences
Period5/07/15 → …


Dive into the research topics of 'Developing an improved sediment-specific biomonitoring tool: Combining expert knowledge and empirical data'. Together they form a unique fingerprint.

Cite this