Determining the quality of taxonomic data
James B. Stribling
Tetra Tech, Inc.
Stephen R. Moulton II
US Geological Survey, National Water-Quality Assessment Program
Gary T. Lester
EcoAnalysts, Inc.
In this article, Stribling et al. discuss data quality issues to be considered when conducting taxonomic analyses for biological assessments. They differentiate between 2 broad areas of taxonomy—research and production taxonomic investigations—and consider how approaches to organism identification can vary between these 2 areas. The authors stress the importance of evaluating and communicating data quality, and that knowledge of quality assurance/quality control elements is essential before drawing conclusions from biological assessment results.
A biological assessment protocol is a measurement system, a series of methods that functions to translate field samples to quantitatively based narrative assessments (Diamond et al. 1996, Barbour et al. 1999). Carter and Resh (2001) showed that there is widespread variability among monitoring programs in field and laboratory methods used in biological assessments. The components of any measurement system can contribute to the variability of the results (Taylor 1988, Warren-Hicks et al. 2000), so it is important for data users to understand the uncertainty associated with them. Therefore, it is necessary to evaluate individual components of the process, because each represents a potential error source (in this sense, we use error and variability interchangeably). The relative importance and acceptability of different error sources depends on specific objectives and data needs (Fig. 1); they typically are stated as measurement or data quality objectives (MQOs and DQOs, respectively) (Costanza et al. 1992, USE-PA 2000).