Assemblage diversity, cell density and within-slide variability: Implications for quality assurance/quality control and uncertainty assessment in diatom-based monitoring. (October 2016)

• Record Type: Journal Article
• Authors: Lavoie, Isabelle; Campeau, Stéphane
• Is Part Of: Ecological indicators. Volume 69(2016)
• Page Start: 415
• Page End: 421
• Publication Date: 2016-10
• Subjects: Assemblage diversity -- Bioassessment -- Bray–Curtis similarity index -- Diatom valve density -- Diatoms -- Hill's N2 -- IDEC -- Within-slide variability; Environmental monitoring -- Periodicals; Environmental management -- Periodicals; Environmental impact analysis -- Periodicals; Environmental risk assessment -- Periodicals; Sustainable development -- Periodicals; 333.71405
• DOI: 10.1016/j.ecolind.2016.05.001
• Languages: English
• ISSNs: 1470-160X

Abstract

Abstract: This study was undertaken to evaluate the variability associated with the microscope analysis step in the application of the Eastern Canadian Diatom Index (IDEC: Indice Diatomées de l'Est du Canada), with the general objective of developing a suitable quality assurance/quality control (QA/QC) program for this biological index. For this purpose, we estimated within-slide variability (replicability) and inter-analysts variability (reproducibility), as a function of diatom assemblage diversity and slide cell density. Overall, our results show that variability associated with diatom assemblage characterization is low, which ensures that IDEC scores reflect environmental changes rather than variability at the microscope analysis step. The main recommendations ensuing from this study are (for the IDEC in particular but also for diatom-based monitoring in general): (1) An error term of ±2 IDEC units corresponding to the within-slide variability (replicability) should accompany all reported IDEC scores. (2) A deviation of ±3 points from the audit's IDEC scores should be considered as an acceptable difference. Considering the above-mentioned estimated error term of ±2 associated with all IDEC scores, an overall deviation of 7 would still be satisfactory. (3) Samples showing low diversity (Hill's N2 ≤5) should automatically be submitted for QA/QC. (4) A Bray–Curtis (analyst vs audit) similarity of ≥60% should also be included as a QA/QC criterion, and should increase to ≥70% for poorly diversified assemblages (Hill's N2 ≤5). (5) A diatom valve density of ≤15 per field of view should be targeted in order to reduce variability at the enumeration step. The results of this study illustrate how a relatively simple and straightforward approach to QA/QC can greatly strengthen the reliability of ecological inferences from an index based on a group of organisms with a high taxonomical diversity. It also highlights the importance of regular communication between analysts in order to maintain a high degree of concordance within taxonomical identification.