1. We describe the dynamics of host–parasite interactions over a period of more than 30 years between the freshwater diatom Asterionella formosa and two highly virulent chytrid parasites (Rhizophydium planktonicum and Zygorhizidium planktonicum) in Lake Maarsseveen, The Netherlands. This period is characterised by a significant warming trend which is strongest in spring.
2. The key spring event in lakes, the diatom bloom, was in many years dominated by Asterionella. We examine whether and how climate warming has affected the prevalence of infection in Asterionella by chytrids.
3. In years with cold winters/early springs, a dense Asterionella bloom is followed by epidemic development of disease as high Asterionella densities greatly facilitate transmission of chytrid zoospores. This sequence of events is absent in milder winters.
4. Earlier experimental studies have shown that the parasite is almost non-infective at water temperatures below 3 °C, offering a disease-free window of opportunity for growth of Asterionella. Climate warming has reduced periods in which water temperature remains <3 °C, narrowing the window of opportunity for uninfected growth. Consequently, Asterionella continuously suffers from infection, albeit at low levels.
5. Population reduction as a result of low level infection allows other diatoms to take over as dominant species, possibly through priority effects.
6. In mild winters, chytrid infections no longer reach epidemic levels, but remain at low prevalence since transmission is impaired at low host densities. Climate warming thus affects both host and parasite in intricate ways, with the host denied a bloom and consequently the parasite denied an epidemic.
7. A shift from Asterionella to a mixed diatom community in years with mild winters may benefit the food web, because of the poor edibility of Asterionella, unless the numerous chytrid zoospores produced during epidemics significantly contribute to zooplankton nutrition.
8. Our study demonstrates the potential complexity of climate change impacts on disease. A reduction in the likelihood of epidemic development of a virulent parasite would seem to be of great benefit to the host, but this was not the case. Unexpected, sometimes paradoxical consequences of climate change can be expected and suggest that the view of a ‘warmer hence sicker world’ may not always apply.