Title:	TRITON – A SYSTEM FOR ANALYSIS OF BIOMONITORING DATA AND ITS ROBUST SOLUTION FOR MULTIVARIATE TYPOLOGY OF RIVER LOCALITIES
Author(s):	Jiří Jarkovský , Petr Pavliš , Ladislav Dušek , Jan Hodovský
ISBN:	972-99353-6-X
Editors:	Nuno Guimarães and Pedro Isaías
Year:	2005
Edition:	2
Keywords:	Robust multivariate analysis, macrozoobenthos, RIVPACS, biomonitoring, Gower metric, information system.
Type:	Short Paper
First Page:	333
Last Page:	338
Language:	English
Cover:
Full Contents:	click to dowload
Paper Abstract:	Monitoring of water organisms communities has become a standard approach in surface water monitoring as well as a part of complex systems for assessing surface water quality. In this paper, we are proposing the TRITON system developed for the analysis of data retrieved from the biomonitoring network of the Czech Republic. The system works in a network environment alongside a database system storing biomonitoring data; it manages rights of the system users and provides analysis and reporting based on this data. In respect of the use of the data on biological communities, several problems emerge connected with the following question to be solved for the TRITON system, i.e. “How to relate biological communities to abiotic properties of the respective localities and water quality classes?” There are several methods solving this problem: simple univariate methods (e.g. biotic indices) or more complex multivariate methods (e.g. RIVPACS or BEAST). In this paper, we are proposing a new point of view concerning water quality assessment – a method based on robust multivariate analysis of macrozoobenthos communities and abiotic properties of the respective sites. It consists of two main components: robust true distances of sites based on several data views (biotic, static and dynamic abiotic properties) and selection of reference groups (i.e. quality classes) of localities. The analysed localities are compared to a reference model using their distances from reference groups’ centroids and probabilistically assigned to quality classes. The TRITON system is now being implemented into the biomonitoring network of the Agricultural Water Management Authority of the Czech Republic and the reference model for multivariate assessment of water quality classes is under development. Other tasks in the future of the TRITON system are i) to establish a system integrating both the abiotic monitoring and biomonitoring data on all types of surface waters in the Czech Republic into one complex system operating on the national level, ii) the biomonitoring programs and their analyses should cover more taxonomical groups of organisms than macrozoobenthos only.
	Go Back