Is it possible to optimise the labour and time intensity of diatom analyses for determination of the Polish diatom indices (IO, IOJ)?

Diatom-based indices are a recommended approach in the assessment of the ecological status of lakes and rivers, required by the EU Water Framework Directive. To optimise the time-consuming microscopic analyses, we determined the lowest number of diatom valves needed to be identified (‘counted’) in a sample to arrive at a reliable ecological status assessment based on the Polish multimetric diatom indices (diatom index, IO for rivers, and diatom index for lakes, IOJ). We compared the indices calculated with different valve counts. The dataset consisted of diatom counts in 55 samples collected in the rivers Ina and Drawa (Western Pomerania, NW, Poland) and their lakes. Following the method used by the Polish Chief Inspectorate for Environmental Protection (ChIEP), periphyton samples were collected from macrophytes growing in the rivers and lakes, and diatom slides were prepared. Originally, the ecological status was determined based on identifying 400 valves per slide. In the subsequent simulation, 50, 100, 150, 200, 250, 300, 350 and 400 valves per slide were randomly drawn 20 times for each sample and each simulation, and the indices were calculated again. In 87% of our simulations, the ecological status turned out to be identical, regardless of the number of valves identified. However, the standard deviation of the index values was at its highest (> 0.05) with 50 valves, and decreased systematically (below 0.05) for 100 and more valves. Simulated datasets showed logarithmic trend of changes where the compilation and standard deviation was stable above 200 valves identified. Therefore, we recommend simplified methodology for Polish multimetric diatom indices. Reduction of identified valve quantity in a sample to 200 valves would result in over 95% of compliance with classification obtained when 400 valves were identified.