While both spatial variation and disease dynamics are two commonly invoked mechanisms for explaining patterns in biodiversity, the interplay of the two is rarely considered. Yet, spatial structure could mitigate the effect of disease on plant community composition, as patterns in autocorrelation, clumping, and dispersal dynamics are known to alter the rate of disease spread. Here, we use the barley yellow dwarf virus (BYDV) system to examine how spatial structure and vector (aphid) dispersal behavior jointly alter California grassland diversity. We conducted bio-assay experiments of vector preference for: (1) healthy versus infected plant tissue and (2) perennial native versus annual exotic grasses to determine differences in vector dispersal preference across healthy and viruliferous vectors. We subsequently created a susceptible-infected ordinary-differential equation model that incorporates dispersal preference in heterogeneous landscapes and parameterized the model from our preference bio-assays. Using this model, we examined how spatial structure and dispersal behavior jointly alter the spread of disease in California grasslands and the corresponding implications for California grassland diversity and invasion dynamics.
Results/Conclusions
In lab trials, vectors exhibited highly biased dispersal behavior where healthy aphids preferred to disperse to infected plants, while viruliferous aphids were more likely to disperse to healthy plants. This is consistent with the vector manipulation hypothesis, where vector behavior is altered by the virus in a manner that promotes disease spread. In a single host model, this preference does, indeed, increase the spread of BYDV in grasslands. In addition, when testing dispersal preference in lab trials in a multi-host context, healthy aphids exhibited no preference for exotic annuals or native perennial plants. Viruliferous aphids, however, exhibited a strong preference to disperse to native perennial plants. Incorporating these results in the differential equation model, surprisingly, slowed the spread of disease, especially in grasslands dominated by perennial grasses, because of the nonrandom distribution of vectors onto given plant types. The diversity of plant hosts therefore plays a significant role in the spread of disease, where increased host diversity slows the spread of BYDV because of biased dispersal. Model results show that spatial structure and biased vector dispersal, in this case, increase the ability for invasion of exotic annual grasses, and correspondingly decrease the abundance of native perennial grasses.