Response to Comment on "Models predict planned phosphorus load reduction will make Lake Erie more toxic".

Huisman et al. claim that our model is poorly supported or contradicted by other studies and the predictions are "seriously flawed." We show their criticism is based on an incomplete selection of evidence, misinterpretation of data, or does not actually refute the model. Like all ecosystem models, our model has simplifications and uncertainties, but it is better than existing approaches hat ignore biology and do not predict toxin concentration.

Fate of hepatotoxin microcystin during infection of cyanobacteria by fungal chytrid parasites.

Chytrid parasites are increasingly recognized as ubiquitous and potent control agents of phytoplankton, including bloom-forming toxigenic cyanobacteria. In order to explore the fate of the cyanobacterial toxin microcystins (MCs) and assess potential upregulation of their production under parasite attack, a laboratory experiment was conducted to evaluate short- and long-term variation in extracellular and intracellular MC in the cyanobacteria Planktothrix agardhii and P. rubescens, both under chytrid infection and in the presence of lysates of previously infected cyanobacteria. MCs release under parasite infection was limited and not different to uninfected cyanobacteria, with extracellular toxin shares never exceeding 10%, substantially below those caused by mechanical lysis induced by a cold-shock. Intracellular MC contents in P. rubescens under infection were not significantly different from uninfected controls, whereas infected P. agardhii showed a 1.5-fold increase in intracellular MC concentrations, but this was detected within the first 48 hours after parasite inoculation and not later, indicating no substantial MC upregulation in cells being infected. The presence of lysates of previously infected cyanobacteria did not elicit higher intracellular MC contents in exposed cyanobacteria, speaking against a putative upregulation of toxin production induced via quorum sensing in response to parasite attack. These results indicate that chytrid epidemics can constitute a bloom decay mechanism that is not accompanied by massive release of toxins into the medium.

Parasites do not adapt to elevated temperature, as evidenced from experimental evolution of a phytoplankton-fungus system.

Global warming is predicted to impact the prevalence and severity of infectious diseases. However, empirical data supporting this statement usually stem from experiments in which parasite fitness and disease outcome are measured directly after temperature increase. This might exclude the possibility of parasite adaptation. To incorporate the adaptive response of parasites into predictions of disease severity in a warmer world, we undertook an experimental evolution assay in which a fungal parasite of phytoplankton was maintained at elevated or control temperatures for six months, corresponding to 100-200 parasite generations. Host cultures were maintained at the respective temperatures and provided as substrate, but were not under parasite pressure. A reciprocal infection experiment conducted after six-month serial passages revealed no evidence of parasite adaptation. In fact, parasite fitness at elevated temperatures was inferior in parasite populations reared at elevated temperatures compared with those maintained under control temperature. However, this effect was reversed after parasites were returned to control temperatures for a few (approx. 10) generations. The absence of parasite adaptation to elevated temperatures suggests that, in phytoplankton-fungus systems, disease outcome under global warming will be largely determined by both host and parasite thermal ecology.

Exposure to nanoplastics affects the outcome of infectious disease in phytoplankton.

Infectious diseases of humans and wildlife are increasing globally but the contribution of novel artificial anthropogenic entities such as nano-sized plastics to disease dynamics remains unknown. Despite mounting evidence for the adverse effects of nanoplastics (NPs) on single organisms, it is unclear whether and how they affect the interaction between species and thereby lead to ecological harm. In order to incorporate the impact of NP pollution into host-parasite-environment interactions captured in the "disease triangle", we evaluated disease outcomes in the presence of polystyrene NP using an ecologically-relevant host-parasite system consisting of a common planktonic cyanobacterium and its fungal parasite. NP at high concentrations formed hetero-aggregates with phytoplankton and inhibited their growth. This coincided with a significant reduction in infection prevalence, highlighting the close interdependency of host and parasite fitness. Lower intensity of infection in the presence of NP indicates that reduced disease transmission results from the parasite's diminished ability to establish new infections as NP formed aggregates around phytoplankton cells. We propose that NP aggregation on the host's surface acts as a physical barrier to infection and, by reducing host light harvesting, may also hamper parasite chemotaxis. These results demonstrate that the consequences of NP pollution go well beyond toxic effects at the individual level and modulate the intensity of species interactions, thereby potentially eliciting diverse cascading effects on ecosystem functioning.