Effects of trophic cascades in dilution grazing experiments: from artificial saturated feeding responses to positive slopes

9 pages, 3 figures, 3 tablesDilution grazing experiments are the most commonly used method for quantifying microzooplankton grazing rates on phytoplankton. However, in some cases, the outcomes of these experiments yield non-significant grazing impacts, even when microbial grazers are numerous, or result in unexpected non-linear functions or positive slopes. Here, we argue that some of these situations and others that may go unnoticed can be explained by trophic cascades occurring in the community confined within the incubation experiments. When trophic cascades are present, simple linear models relating dilution levels to phytoplankton net growth rates can be distorted into artificial saturation curves, V-shaped responses or a non-significant fit and may even invert the sign of the slope. These artefacts have minor consequences for true estimates of phytoplankton mortality rates when the dilution level is sufficiently high (ca. 5-10%), but they may result in significant underestimation of the microzooplankton grazing impact when working at lower, typical maximum dilution levels (e.g. 25%). These underestimations can be particularly important when the microzooplankton community is already under feeding-saturated conditions. © The Author 2013This research was funded by the PROTOS project (CTM2009-08783) of the Spanish Ministry of Science and Innovation assigned to A.C., and it is a contribution of the Marine Zooplankton Ecology Excellence Group from the Generalitat de Catalunya (2009SGR-1283)Peer Reviewe