Fossil fuel combustion and agricultural fertilization have dramatically increased global pools of nitrogen (N) and phosphorus (P) relative to pre-industrial levels.  Fertilization can control tissue chemistry in individual plants via luxury uptake and storage, as well as changing the competitive balance among species in a community.  Thus, this widespread fertilization of terrestrial communities could have important direct effects on plant composition, as well as indirectly controlling herbivory rates or pathogen transmission via insect-vectored pathogens.  Barley yellow dwarf virus (BYDV), a globally-distributed generalist pathogen of all Poaceae, is obligately transmitted by aphids.  In lab experiments examining the fecundity of aphids in response to a factorial addition of N and P we found that N fertilization increased % tissue N and aphid fecundity, whereas P fertilization did not affect either of these response variables.  Movement/emigration of aphids increases with aphid density on individual plants which can lead to increased pathogen transmission, so our lab results led us to hypothesize that fertilization should result in increased pathogen prevalence in a field setting. In 100m² plots in an annual grass-dominated southern California grassland, we added a factorial combination of N, P, potassium (K) and micronutrients and measured the prevalence of BYDV infection in a common and widespread annual grass, Bromus hordeaceus.
Results/Conclusions

We found that, contrary to our lab predictions, BYDV prevalence declined with increasing N.  Importantly, prevalence declined with declining % cover of host species and increasing % cover of non-hosts (mostly forbs) across the N-addition gradient.  Thus, N-fertilization appears to determine host infection primarily by controlling the host community context – competitive outcome between grasses and forbs – rather than by aphid responses to changes in individual host tissue chemistry.