Diatoms and biomonitoring: should cell size be accounted for?

Despite the fact that biovolume calculation is a common procedure in most phytoplankton and periphyton studies, diatom community analyses are usually based on relative abundance data. In a biomonitoring context, a community metric that accounts for cell size could be of interest due to the potential differences that might exist in nutrient uptake between large and small-sized species. This paper addresses the question of whether diatom community analysis should be based on relative abundance, biovolume or cell surface. The results show that although community structure expressed as relative proportion of taxa varied according to the metric used, the ordinations conducted with each metric were similar. The explained percentage of species variance was slightly higher with the relative abundance metric compared to the metrics based on relative biovolume or cell surface area. Partial CCAs showed that each water chemistry variable generally explained a higher portion of species variance when the relative abundance was used. The analyses conducted with two size groups (small and large taxa) expressed as relative abundance and relative biovolume showed similar results. Moreover, our data showed that there is no significant relationship between diatom size and total phosphorus. According to these results, it seems that relative abundance would be the most appropriate metric to use for biomonitoring purposes. The biovolume and cell surface area calculations added substantially to the total analysis time due to the numerous measurements required, but did not improve the variance explained in community structure, and site ordinations were not significantly different.