Phytoplankton precision trials in the enumeration and identification of marine microalgae through the scheme “Biological Effects Quality Assurance in Monitoring Programmes (BEQUALM)”.

Abstract

Scientists are coming under increased pressure in recent years to show that results they obtain arising from their scientific work are quality assured and stand up to scrutiny by independent expert auditors. This has meant that the methodologies used by laboratories involved in making these measurements have to be validated and fit for purpose and has led to the adoption of internationally recognised standard protocols. These protocols must be underpinned by robust quality systems and must be accredited to an international standard. In order for laboratories to become accredited in particular methods, they have to fulfil a series of prerequisites but a compulsory one is the participation in a proficiency testing scheme. Proficiency testing schemes are independent assessor organisations which coordinate regular inter-calibration and intercomparative studies between laboratories with a common purpose. What happens, though when proficiency testing schemes do not exist for a particular scientific measurement? This study presents results from two inter-comparison exercises at European level between phytoplankton monitoring laboratories in the enumeration and identification of marine microalgae using the Utermöhl cell counting method. Microalgae are a very important ecological component of the marine ecosystem and have also become important ecological indicators of hydro-climatic change, ocean acidification and eutrophication. Member states of the European Union are obliged to monitor for toxic and harmful algae which can cause problems and devastation in the natural environment, have detrimental effects on human health if contaminated fish and shellfish are eaten, can cause huge economic losses to the aquaculture industry and impact directly in coastal communities. This study shows how an intercomparison of this kind is designed and organised, how samples are set up, materials homogenised and reference values obtained. It demonstrates the importance of using the right technique and best practice, based on experience, to analyse samples and how important it is to design the exercise to be statistically robust, both quantitatively and qualitatively. The taxonomy quiz in 2009 showed that there was no evidence that video clips were better audit trail tools than images or vice versa. The quantitative measure suggested that there was evidence of good agreement between virtually all the analysts and the reference value for all species except one (P.micans). However, there was evidence of lack of reproducibility between and within laboratories. The qualitative measure calculated indicated that analysts are more likely to identify a toxic organism as a non- toxic organism than the other way around. The results from the enumeration data in the 2010 exercise showed that there was lack of reproducibility across laboratories using different counting strategies and volume sub-sampled and analysed. These results when compared to a set of hypothesised means used as reference values suggested that cell counts were potentially underestimated by as much as 30% and that this underestimation was most likely due to test method effects.