Summary

1. Landscape composition affects local arthropod biodiversity, including herbivorous insects and their predators, yet to date landscape effects on insect-vectored plant diseases have received little attention. Here, we examine how landscape composition affects the prevalence of a viral pathogen in host plants, and the role the arthropod vector assemblage plays in mediating landscape effects.

2. We measured the effect of landscape composition (measured as percentage of cropland and unmanaged land) on the plant virus Potato virus Y (PVY), its aphid vectors, and their coccinellid predators during the 2012 and 2013 field seasons at 19–21 farms.

3. In both years, we found a positive relationship be tween final virus prevalence and percent- age of cropland within 500, 1000 and 1500 m surroundi ng study sites. Percentage of c ropland also had a significant negative effect on aphid species richness, and the aphid community composition in turn affected PVY prevalence. By contrast, landscape composi tion had no measurable effect on coccinel lid abundance or species richness in this study.

4. Synthesis and applications. Our work demonstrates that landscape composition plays an important role in vector-borne pathogen spread, and that pathogen spread appears to be mediated by the effects of the landscape on the insect vector community. The small spatial scale (≤1500 m) of the effects seen in our study indicates that on-farm management practices have the potential to reduce virus prevalence on small-scale farms. Farmers may be able to reduce Potato virus Y prevalence by on-farm diversification, by isolating potato fields from other a gricultural crops, and by not using saved potato seed.