Trophic strategy determines stoichiometry of plankton. In general, heterotrophic zooplankton have lower and more stable C:N and C:P ratios than photoautotrophic phytoplankton whereas mixotrophic protists, which consume prey and photosynthesize, have stoichiometry between zooplankton and phytoplankton. As trophic strategies change with cell size, body size may be a key trait influencing eukaryotic plankton stoichiometry. However, the relationship between body size and stoichiometry remains unclear. Here, we measured plankton size-fractionated C:N ratios under different intensities of light and nutrient supply in subtropical freshwater and marine systems. We found a unimodal body size-C:N ratio pattern with a maximum C:N ratio at ~50 μm diameter in marine and freshwater systems. Moreover, the variation in C:N ratios is mainly explained by body size, followed by light intensity and nutrient concentration. To investigate the mechanisms behind this unimodal pattern, we constructed a size-based plankton food web model in which the trophic strategy and C:N ratio is an emergent result. Our model simulations reproduce the unimodal pattern with C:N ratio of photoautotrophs ≤ 50 μm increasing with body size due to increase of photosynthetic carbon, whereas C:N ratios of organisms > 50 μm decreases with size due to decreasing photoautotrophic but increasing heterotrophic uptake. Based on our field observations and simulation, we extend the classic “light-nutrient” theory that determines plankton C:N ratio to include body size and trophic strategy dependency. We conclude that body size and size-dependent uptake of resources (light, nutrients and prey) determine plankton stoichiometry at various light and nutrient supplies.

The datasets of freshwater and marine size-fractionated plankton C:N ratio were collected by Norpac net of 50 and 200 um in Feitsui Reservoir and East China Sea. The plankton collected were seperated to several size classes (see Appendix A of the manuscript) and the C and N contents were measured by elemental analyzer (EA1108, Fisons, Italy, and FLASH 2000, Thermo SCIENTIFIC, USA). The molar C:N ratios were calculated according to the C and N contents of the plankton.

For model simulation, the original m-files can be found here. Parametere_AN.m includes the parameters that were kept constant in each simulation. ControlParam_AN.m includes parameters that could be changed in the simulation, such as the nutrient supply and light intensity. MixotrophWeb_AN.m is the main model ODEs.

Missing data in Table S1 and S2 are labeled NA.