Fear your tolerant neighbors: Host defense and vector preference determine disease spillover in an epidemiology model

Background/Question/Methods

Variation in host resistance and tolerance—collectively defined as defense—to pathogens can strongly influence the co-evolution of both host and pathogen.  However, the effects of plant defense on disease spread are not well understood, particularly for vector-borne pathogens.  Vector discrimination between healthy (asymptomatic) and diseased (symptomatic) hosts has long been recognized for its importance for disease spread.  However, previous authors have assumed a host population that is uniformly susceptible to the pathogen. Here we relax that assumption to investigate how vector preference interacts with host resistance or tolerance.

We developed a vectored-SEI (Susceptible, Exposed, Infective) epidemiology model, with vector preference for either diseased or healthy hosts.  Plant tolerance to infection was varied by incorporating a Carrier state into the model; hosts that are pathogen sources but show no symptoms.  Plant resistance was varied by modifying the incubation rate at which Exposed individuals became either Infectives or Carriers.  Finally, we compared disease dynamics in one-patch and two-patch models.  In the two-patch model, we explored the potential for disease spread between host patches with varying host defenses.

Results/Conclusions

Analysis of the single-patch model showed that lower resistance caused greater disease spread, and that this effect was largely independent of vector preference.  In contrast, the effect of plant tolerance on disease spread was highly sensitive to vector preference.  Preference for healthy hosts within tolerant host populations caused high levels of disease spread, because vectors could still acquire the disease from tolerant hosts, even though the hosts appeared to be disease-free. 

We found qualitatively similar effects between the one- and two-patch models.  When vectors avoid diseased hosts, the greatest disease spillover into undefended patches was from neighboring host patches with high tolerance.  In contrast, when vectors prefer diseased hosts, the greatest disease spillover into undefended patches was from neighboring susceptible host patches.  Our results indicate complex, interacting effects between host defense and vector preference for disease spread.  This work highlights the ecological importance of variation in host defense against pathogens, and has important implications for crop and livestock breeding programs for disease management.