Plant host physiology and risk of infection with generalist, vector-borne pathogens

Background/Question/Methods

Several approaches exist to predict the transmission of generalist pathogens.  These include the concept of the disease triangle, ideas as to how changing host community diversity and/or composition can influence transmission, and mathematical models of transmission.  These concepts and methods are useful predictors of pathogen dynamics in specific systems, but, because they have yet to be integrated into a larger body of ecological theory, they are not broadly predictive.  That is, for each new system, models need to be redesigned and parameterized, important reservoir hosts identified, and environmental influences determined. In this talk, we will explore the potential for host phenotype, a multi-dimensional metric of host physiology, to integrate host genetic and environmental effects on host-vector and host-pathogen interactions. A recent but limited test of this hypothesis suggested that plant host phenotype can explain variation in host susceptibility to generalist, vector-borne pathogen infection and host contribution to infected vector populations.  If this pattern is general, host phenotype could be used to inform models of transmission and to identify hosts with high reservoir potential.  Is this pattern general across a wide suite of host species and environmental gradients?  If so, can we scale up and use host phenotypes to predict transmission in multi-host communities of varying diversity and composition?  Can phenotypes be used to predict infection risk to focal hosts of interest? 

Results/Conclusions

We present the most comprehensive test to date of the influence of host physiological phenotype on epidemiology.  In the greenhouse, we manipulated 22 host species and nitrogen availability.  Across species and nitrogen treatments, host phenotype was defined by a principal component axis of increasing tissue nitrogen, photosynthetic capacity, and growth rate.  Phenotype is a strong predictor of host susceptibility to a generalist, vector-borne pathogen (Barley yellow dwarf virus-PAV), but not as strong as host lifespan.  This suggests that we have identified some mechanisms underlying the influence of host lifespan on epidemiology, but others remain to be identified.  We will suggest a method for scaling up from individual- to community-level host phenotype in order to predict transmission in host communities of varying diversity and composition.  We also suggest that infection risk to a focal host is determined by an interaction between focal host and community-level phenotypes.  By using host phenotype to explain host-pathogen interactions, we aim to help integrate disease ecology into broader ecological theory, particularly the mechanistic concepts of organism-environment interactions from physiological ecology.