Interactions among parasites and mutualists may depend on priority effects both across host populations due to seasonal variation and within host individuals due to differences in force-of-infection during epidemics. We tested the roles of seasonal variation in epidemic timing, within-host priority effects, and other direct interactions on symbiont community assembly by measuring parasite infection on three cohorts of sentinel tall-fescue plants, timed to detect seasonal dynamics, and grown with and without the fungal mutualist, Epichloë coenophiala. We focused on three foliar fungal parasites (Colletotrichum cereale, Puccinia coronata, and Rhizoctonia solani) that exhibit overlapping seasonal epidemics. The first cohort was placed during the Colletotrichum epidemic before other parasite epidemics began. The second cohort was placed shortly after the Rhizoctonia epidemic began, before Puccinia emerged. The third cohort was placed shortly after the Puccinia epidemic began, when all three focal parasites were present. To evaluate the role of within-host priority effects, we performed time-until event analysis (Cox-Proportional Hazards), specifically testing whether previous infection status by each symbiont influenced the risk of subsequent infection by each parasite (i.e. disease risk). To test the role of direct interactions among infected hosts, we analyzed infection severity, which measures within-host parasite replication (i.e. growth rate).
Results/Conclusions
Seasonal variation, within-host priority effects, and direct interactions influenced symbiont assembly together. For example, the focal parasite, Colletotrichum, infected 75% of host leaves in the first cohort, but never exceeded 25% infection prevalence in later cohorts. Despite benefiting from a seasonal effect in the first cohort, Colletotrichum disease risk and growth still responded strongly to other symbionts. On average, disease risk was 6x higher among leaves that were previously infected with Puccinia (p<0.01) and 1.5x lower on leaves previously infected with Rhizoctonia (p<0.05). Colletotrichum also exhibited slower growth in endophyte-infected hosts (p<0.05) and in leaves with more severe Rhizoctonia infections (p<0.05). In the second and third cohorts, where Colletotrichum had no seasonal priority, Rhizoctonia prevalence increased from < 25% to > 75%. Yet, in all cohorts, Rhizoctonia still experienced significant within-host priority effects and direct interactions with other symbionts. In contrast, Puccinia was strongly regulated by within-host Rhizoctonia priority effects (p<0.001) and experienced highest prevalence when Colletotrichum suppressed Rhizoctonia. Together, these results indicate that strong seasonal priority effects do not necessarily preempt within-host priority effects, which do not necessarily preempt direct interactions between symbionts. Instead priority effects and direct interactions likely act in concert during symbiont community assembly.