Skip to main content


Title: Lake Benthic Macroinvertebrates 2: Quantifying Uncertainty in Sampling Methodology
Author: I.J. Jones
Author: R.T. Clarke
Author: J.H. Blackburn
Author: R.J.M. Gunn
Author: N.T. Kneebone
Author: M.W. Neale
Author: Environment Agency
Document Type: Monograph
Annotation: Environment Agency Project ID:EAPRJOUT_1621, Representation ID: 559, Object ID: 2729
While the first report in this series (CEH, 2006) reviewed existing sampling methodology and resulted in a series of recommendations, this second report specifically assesses the causes and levels of uncertainty inherent within the revised method. This work is important since the methods used for assigning an individual water body to an ecological status class are based on sample information and therefore subject to uncertainty. Specifically, the method relies on field sampling to estimate the values of individual EQRs and then using agreed rules to combine the information from different metrics into an overall assessment for the water body. Given the nature of sampling, this means that other replicate samples from the same water body at the same time could give rise to different values of one or more Ecological Quality Ratios (EQRs) and hence, possibly, even different estimates of the water body’s status. The aim of this study was, therefore, to assess the variation and uncertainty associated with a selected lake benthic macro-invertebrate sampling method, the ultimate goal being to assist in those tasked with developing tools that provide a sufficiently robust classification for assessing the ecological status of lakes. This study demonstrated that community composition of stations within a lake were, on average, always more similar to other stations from the same lake, than to stations from a different lake. It also revealed significant differences in metric values were found between lakes within a pressure class, contributing 83%, 74%, 29%, 82%, 85% and 74% of the total variance of BMWP families, BMWP score, ASPT, AWIC families, AWIC total score and AWIC, respectively. Finally, this study was able to demonstrate the effect of sorting or identification errors at family level on the variation. Sorting errors were responsible for 6-11% of the total variance, sample-processing errors causing an overall tendency to under-estimate the ‘true’ value of AWIC for a sample, but not ASPT. However, the effects of variation due to field sampling, sorting and identification errors can be incorporated into estimates of the uncertainty in bioassessments using the software package STARBUGS (STAR Bioassessment Uncertainty Guidance Software).
Publisher: Environment Agency
Subject Keywords: Water Framework DirectiveLakesRIVPACSLittoral zoneInvertebratesSamplingBenthic environmentLoughsLochs
Geographic Keywords: United Nationscertainty
Extent: 45
Total file downloads: 292

Download PDF    Display PDF in separate tab