Skip to main content

PDF


Title: Phytoplankton Classification tool for UK Lakes
Author: Carvalho L.
Author: I. Dodkins
Author: B. Dudley
Author: S. Maberly
Author: SNIFFER
Document Type: Monograph
Annotation: Environment Agency Project ID:EAPRJOUT_1347, Representation ID: 451, Object ID: 2522
Abstract:
WFD38: Phytoplankton Classification Tool for UK Lakes: Phytoplankton Composition (October, 2006) Project funders/partners: SNIFFER and Environment Agency Background to research The Environment Agency and SNIFFER have commissioned this R and D project to develop a method to classify the ecological status of lakes on the basis of phytoplankton. As part of this assessment, metrics need to be developed for phytoplankton community composition. Objectives of research Specific objectives for the project were to develop a robust classification, incorporating: 1. Prediction of reference scores for UK lakes based on phytoplankton composition 2. Developing criteria for defining the good/moderate boundary 3. Classifying the ecological status of a water body in to one of five status classes (High/Good/Moderate/Poor/Bad), based on the calculation of an Ecological Quality Ratio (EQR). An EQR being calculated from the relationship between current observed and reference phytoplankton community composition for a site 4. Determining uncertainty associated with the classification result, based on statistical confidence or probability of class Key findings and recommendations Following collation of a dataset of matching phytoplankton and environmental data from 189 lake samples, a multivariate approach to metric development was adopted. CCA was used to develop a species-environment model for phytoplankton, with the main typology variables (alkalinity, altitude, mean depth, lake area) included as significant explanatory variables in the model alongside two variables indicative of eutrophication pressure (chlorophyll and TP concentrations). The model indicated strong correlations between a number of the explanatory variables, with the eutrophication pressure gradients (Chlorophyll and TP) closely correlated with alkalinity. This highlighted the potential problem of developing simple univariate optima of phytoplankton taxa against pressure gradients. The current model explained only 6.8% of the variance in the phytoplankton composition data. This is low, but fairly typical of ecological datasets with large numbers of taxa with greatly varying biomass. The use of abundance data improved the model performance a little, but may need to be re-considered after estimates of analysis error of biovolume measurements have been quantified. A simpler model using presence/absence data may be acceptable. Further enlargement of the dataset, addition of further explanatory variables, such as colour or flushing rate, and taxonomic harmonisation should all help improve the model. Optima were derived for 66 of the most common phytoplankton genera along both eutrophication gradients (chlorophyll and TP) using reciprocal averaging. Although this was still a univariate approach, the correlative effect with alkalinity (and to a lesser extent other typology variables) was removed through the calculation of an EQR by taking account of a siteas typology in the reference score. This was done through the development of a regression model relating reference site scores to typology variables. i Internal validation of the derived metric showed a fairly strong correlation with chlorophyll concentrations in the water column (r2 = 0.53). External validation on an independent dataset is, however, required to more accurately reflect the strength of the relationship. Currently to obtain an EQR between 0 and 1 involves two transformation steps. Further guidance from LTT or ECOSTAT as to what forms of EQR scaling and transformation are acceptable would be beneficial. Currently the H/G boundary is determined from the 75% of reference site scores, giving an EQR of 0.65. The remaining boundaries were derived from an equal division of the EQR scale between 0 and 0.65. This does, however, assume that the maximum impact score observed represents bad status with an EQR of 0. Data gaps and Further Work There is great scope for improving the phytoplankton model through further data collection. Despite 380 phytoplankton samples being counted, only 189 samples had matching chemistry and typology data, with Scottish samples (and many reference lakes) having a particularly poor match. Further data collection from all reference lake types are required (particularly shallow and deep medium and high alkalinity lakes). Across the whole pressure gradient, further data are required from very shallow low and medium alkalinity lakes and deep medium and high alkalinity lakes. Currently the nutrient pressure gradient is not spanned evenly with particular lake types either having few reference sites (e.g. high alkalinity lakes) or few highly impacted sites (e.g. low alkalinity lakes). Currently no data from Northern Ireland and the Republic of Ireland are included in the model. Phytoplankton samples from these regions need to be counted following the project standard guidelines and taxonomy and need to incorporate biovolume measurements. Additional environmental data also needs recording and collating. Mean depth and alkalinity data are needed from all sampled sites and additional data on colour (Hazen units) and (modelled) flushing rate would be beneficial to examine if they added additional, independent, explanatory power to the model. As well as model and metric development, the possibility of identifying class boundaries based on ecological thresholds (e.g. ratio of positive to negative indicators) needs to be examined. Further work is also required on estimates of error and consequent uncertainties in classification. Errors associated with sampling variability (by location and season) and analytical (counter) error requires further data collection. Development of a standard harmonised list of commonly recorded taxa would be of benefit alongside regular (annual) taxonomic workshops to minimise counter error. Key words: phytoplankton, WFD, classification, lake, ecological status ii SNIFFER WFD38: Phytoplankton Classification Tool October, 2006 List of Contributors Laurence Carvalho Centre for Ecology and Hydrology Bush Estate Penicuik Midlothian EH26 0QB Email: laca@ceh.ac.uk Ian Dodkins University of Ulster Coleraine Fiona Carse and Stephen Maberly Centre for Ecology and Hydrology Lancaster Bernard Dudley Centre for Ecology and Hydrology Edinburgh Sian Davies Environment Agency Haddiscoe Acknowledgements The following phytoplankton counters provided the data essential for tool development and also offered advice on developing counting guidance: Jo Girvan, Jane Fisher, Nadia Solvieva, Genevieve Madgwick, Sarah Pritchard, Bill Brierly and Tom Barker The authors would also like to thank Nigel Willby and Robert Ptacnik for discussions on methodological approaches iii SNIFFER WFD38: Phytoplankton Classification Tool October, 2006 TABLE OF CONTENTS EXECUTIVE SUMMARY LIST OF CONTRIBUTORS III ACKNOWLEDGEMENTS III 1.
Publisher: Environment Agency
Subject Keywords: Phytoplankton; Classification; Lake; Wfd; Ecological status
Extent: 43
Permalink: http://www.environmentdata.org/archive/ealit:4739
Total file downloads: 4

Download PDF    Display PDF in separate tab