Interactive Effects of Temperature and Available Hosts on Infection
Among live D. magna treatments, the more resistant clones were less likely to become infected and had a longer time to infection than the more susceptible ones. Moreover, host resistotype interacted with temperature to influence infection prevalence, with most observed infections at low temperature being among the more susceptible clones. As our broadly resistant clones are susceptible to at least one of the known P. ramosa strains in the spore bank, it is not surprising that they still became infected. However, that fewer individuals did so compared to the more susceptible clones (especially at low temperatures) and that they also took longer to become infected could be due either to the frequency of P. ramosa strains within the mud that are able to infect them, or the possibility that these strains indeed take longer to cause visible infection (i.e. reproduce inside the host slower than other strains). Although we are unable to elucidate which of these two hypotheses is more likely with these data, these findings nevertheless suggest an important interactive effect of host resistance with temperature on outbreak timing and dynamics. Similar temperature-by-genotype interactions have been found in D. magnaresistance to other parasites (Bruijning et al. 2022, Santos and Ebert 2022) and in many other host-parasite systems (Judelson and Michelmore 1992, Gsell et al. 2013). Our results imply that the host-parasite relationships in this system are strongly linked to temperature, and that changes in temperature may alter the infection and host demographic cycles we have repeatedly observed.
Prevalence in the ephippia treatments at the end of the experiment was at all temperatures considerably lower than in the treatments with liveD. magna (compare Figure 2c,d). Moreover, within ephippia treatments no infection was observed below 15 oC a finding consistent with dynamics observed at the field site (Ameline et al. 2020). Although we do not know the resistotypes of animals used in this treatment, it is unlikely that this result is due to resistotype composition, as the known resistotypes used in the live treatment are also known to hatch from ephippia (Ameline et al. 2020, Ameline et al. 2021). It is possible that females hatching from resting stages behave differently than asexually produced females and that this reduces the likelihood of encountering the parasite. Sediment-borne parasite stages, like the spores of P. ramosa , are picked up when the host browses over the sediment surface, to enrich its food (Arbore et al. 2016). We do not know if this behaviour differs among the two types of females, but if so, it might partially explain these observed differences and possibly why Spring epidemics of P. ramosa are delayed, because all D. magna in early Spring hatch from ephippia. This hypothesis could further explain why infections are not observed in nature at lower temperatures, since the time for the first generation of asexual females to be born compounded with the impacts of temperature on development would further prolong the time until outbreak.