Synergistic effects of predation and parasitism on competition between edible and inedible phytoplankton

Fungi can affect aquatic ecosystems through syntrophic and parasitic interactions with other organisms and organic matter. In pelagic systems, fungal parasites on phytoplankton can control trophic interactions and food-web dynamics, e.g., zooplankton grazing on fungal parasite zoospores creates an alternative energy pathway (termed “mycoloop”) from otherwise inedible phytoplankton species. We aim to investigate how the mycoloop influences community dynamics in aquatic food-webs combining experimental and modelling approaches. We assembled an experimental system consisting of an inedible (host) phytoplankton species and its parasitic chytrid, an edible (non-host) phytoplankton species, and a zooplankton grazer. Chytrids parasitizing increased edible phytoplankton abundance, while zooplankton grazing decreased edible phytoplankton abundance. In the presence of zooplankton and chytrids, competition effects between edible and inedible phytoplankton species depended on nutrient levels. At high nutrient levels, competition was balanced by an indirect positive chytrid effect and negative zooplankton grazing effects on edible phytoplankton. In contrast, at low nutrient levels, we found chytrid had a negative impact on edible phytoplankton synergistically with zooplankton. Mathematical investigations suggest that the synergistic effect can be caused by the mycoloop. This indicates that the mycoloop substantially affects predator-prey interactions and phytoplankton competition with yet unknown ecological consequences.

where Daphnia was able to suppress the edible Cryptomonas. high Daphnia density was not observed. Moreover, we observed additional patterns for the same 1 4 0 experimental setup as the predation food-web under high-nutrient conditions ( Fig. 1h and Fig. 1j). In 1 4 1 addition to the weak impact on the predation-food-web, Daphnia could not suppress edible Cryptomonas 1 4 2 ( Fig. 1j). Yet, average abundance of infected inedible Staurastrum in the weak impact case was higher 1 4 3 than in the strong impact case. To investigating the combined effect of predation and parasitism, we 1 4 4 mainly focused on the "strong impact" (Fig. 1h) pattern. In the strong impact case, infection levels of 1 4 5 chytrids were low, and density of Daphnia was lower than that under low nutrient conditions for the 1 4 6 entire food-web ( Fig. d and h).

4 7
To investigate the influence of community structure on phytoplankton competition, we 1 4 8 visualized the dynamics of the ratio of edible Cryptomonas to total cell densities of phytoplankton over results, we focused on model and experimental competition results on day 20 because for both high and 1 7 0 low nutrient conditions, differences between the food-web systems were most pronounced between days 1 7 1 15-25 (Fig. 2). To accomplish this, we classified the competition results of systems in the model based on 1 7 2 the ratio of edible Cryptomonas to the total phytoplankton density, as in the experiment (see Methods).

7 3
We prepared two distinct food-web configurations for the entire food-web to investigate the 1 7 4 effect of the mycoloop on the competition results. For this, we implemented models with the entire food-1 7 5 web (i) in presence versus (ii) in absence of the mycoloop (no zooplankton ingestion of chytrids). As 1 7 6 several physiological parameters remain unknown, we investigated the changes in system dynamics 1 7 7 varying specific parameters within biologically reasonable ranges and investigating the results using 1 7 8 phase diagrams for all possible pairwise parameter combinations (Fig. 3, Fig. S3-S6). These results 1 7 9 confirm that the presence of the mycoloop likely results in a synergistic effect (Fig. 3). This effect is not  physiological parameters (Fig. 3), e.g. they show that zooplankton and chytrids additively or 1 9 0 synergistically influence phytoplankton competition results. The laboratory experiment showed Our 1 9 1 model analysis supports that the effects can be explained by the mycoloop. In the performed laboratory 1 9 2 experiment, edible Cryptomonas had a higher relative concentration within the infection-food-web than in 1 9 3 the predation-food-web (Fig. 2), independent of nutrient conditions. This reflects the direct negative 1 9 4 impact of Daphnia on Cryptomonas by predation, while parasites indirectly have a positive impact on 1 9 5 Cryptomonas by infecting their competitors (i.e., inedible Staurastrum). The relative abundance of 1 9 6 Cryptomonas was higher in the infection-food-web than in the predation-food-web under both nutrient 3). Thus, the model potentially explains the difference between the synergistic and additive effects from 2 2 4 the experimental observations by comparing the model results between high-and low-nutrient conditions 6 8 embracing species diversity and multiple interaction types in a framework that allows for adaptive 2 6 9 dynamics to assess long-term trends.

7 0
Recent studies have revealed that next to predators, parasites can have a large effect on 2 7 1 aquatic community stability and dynamics, and have highlighted the need for data to assess the impact of To test whether Daphnia magna consumes the inedible phytoplankton Staurastrum sp., we conducted a 2 9 2 preparatory laboratory experiment. We inoculated the inedible algae Staurastrum without (control) and 2 9 3 with two starved Daphnia magna into a 50 ml tube containing modified CHU-10 medium (hereafter 2 9 4 mCHU-10). The initial density of Staurastrum was adjusted to 1,000 cells ml -1 . After inoculation, the The host-parasite community model system consisted of the host Staurastrum sp., (strain STAU1), and 3 0 5 the parasite Staurastromyces oculus (strain STAU-CHY3), which were established in a previous study [5].

1 0
For these experiments, we used a semi-batch culture system (Fig. S7). At the beginning of the sampled Daphnia to the flask to avoid the extinction of Daphnia by a stochastic process of demography.

2 3
To monitor phytoplankton cell density, we used image-based flow cytometry (FlowCam CYANO, 8000 The food-web model reflected      Table S2.