Grapevine leafroll disease (GLD) is caused by a complex of vector-borne virus species in the family Closteroviridae. GLD is present in all grape-growing regions of the world, primarily affecting wine grape varieties. The disease has emerged in the last two decades as one of the major factors affecting grape fruit quality, leading to research efforts aimed at reducing its economic impact. Most research has focused on the pathogens themselves, such as improved detection protocols, with limited work directed toward disease ecology and the development of management practices. Here we discuss the ecology and management of GLD, focusing primarily on Grapevine leafroll-associated virus 3, the most important virus species within the complex. We contextualize research done on this system within an ecological framework that forms the backbone of the discussion regarding current and potential GLD management strategies. To reach this goal, we introduce various aspects of GLD biology and ecology, followed by disease management case studies from four different countries and continents (South Africa, New Zealand, California-USA, and France). We review ongoing regional efforts that serve as models for improved strategies to control this economically important and worldwide disease, highlighting scientific gaps that must be filled for the development of knowledge-based sustainable GLD management practices.

The grasses Sorghastrum nutans (Indian grass), Schizachyrium scoparium (little bluestem), Panicum virgatum (switchgrass), and Andropogon gerardii (big bluestem) are four of the most common plant species present in a tallgrass prairie (1). Infection with barley yellow dwarf virus (BYDV, family Luteoviridae) is of interest in these species because of the potential effects of the virus on tallgrass prairie plant communities and the potential for tallgrass prairie to function as a reservoir of the virus for infection in wheat or barley fields. In a previous inoculation experiment, an unidentified strain of BYDV transmitted by the aphid species Rhopalosiphum padi was reported to infect S. scoparium but none of the other three grass species (2). We sampled for the presence of five virus strains in at least 50 blooming plants of each grass species in a natural tallgrass prairie stand in August 2000. Samples were collected in watersheds that were designated 1B, 1D, K1A, 20B, and 20C at Konza Prairie Biological Station in the Flint Hills near Manhattan, KS. To detect the virus, we used enzyme-linked immunosorbent assay (ELISA) with antibodies purchased from Agdia (Elkhart, IN). For the PAV, MAV, RMV, and SGV strains, we used double-antibody sandwich ELISA with alkaline phosphatase label. For Cereal yellow dwarf virus (RPV), we used compound direct ELISA with alkaline phosphatase label. The scoring of ELISA results was based on comparison with infected and uninfected control plants of the same species. Symptoms of infection in the field were difficult to interpret visually, since plants in this natural environment often showed multiple symptoms of stress. None of the five strains were detected in 51 individuals of S. nutans. For 50 individuals of S. scoparium, the incidence of infection by the different strains was 4% for MAV, 0% for PAV, 2% for RMV, 0% for RPV, and 58% for SGV. For 51 individuals of P. virgatum, the incidence of infection was 31% for MAV, 0% for PAV, 0% for RMV, 0% for RPV, and 4% for SGV. For 64 individuals of A. gerardii, the incidence of infection was 59% for MAV, 0% for PAV, 0% for RMV, 0% for RPV, and 3% for SGV. The impact of BYDV on these tallgrass prairie species

The enemy-release hypothesis (ERH) states that species become more successful in their introduced range than in their native range because they leave behind natural enemies in their native range and are thus "released" from enemy pressures in their introduced range. The ERH is popularly cited to explain the invasive properties of many species and is the underpinning of biological control. We tested the prediction that plant populations are more strongly regulated by natural enemies (herbivores and pathogens) in their native range than in their introduced range with enemy-removal experiments using pesticides. These experiments were replicated at multiple sites in both the native and invaded ranges of the grass Brachypodium sylvaticum. In support of the ERH, enemies consistently regulated populations in the native range. There were more tillers and more seeds produced in treated vs. untreated plots in the native range, and few seedlings survived in the native range. Contrary to the ERH, total measured leaf damage was similar in both ranges, though the enemies that caused it differed. There was more damage by generalist mollusks and pathogens in the native range, and more damage by generalist insect herbivores in the invaded range. Demographic analysis showed that population growth rates were lower in the native range than in the invaded range, and that sexually produced seedlings constituted a smaller fraction of the total in the native range. Our removal experiment showed that enemies regulate plant populations in their native range and suggest that generalist enemies, not just specialists, are important for population regulation.

Coinfections are common, leading to pathogen interactions during transmission and establishment in a host. However, few studies have tested the relative strengths of pathogen interactions in vectors and hosts that determine the outcome of infection. We tested interactions between two genetically distinct variants of the mealybug-transmitted Grapevine leafroll-associated virus 3. The transmission efficiency of each variant in single variant inoculations by two vector species was determined. The effects of vector species, a coinfected source, and simultaneous inoculation from multiple hosts to one host on variant establishment were examined. Within-vector interactions could have a role in transmission from hosts containing mixed infections, but not when vectors were moved from separate singly infected source plants to a single recipient plant. The invasive Planococcus ficus (Signoret) was a more efficient vector than Pseudococcus viburni (Signoret). Transmission efficiency of the two variants did not differ in single variant inoculations. Overall infections were the same whether from singly or coinfected source plants. In mixed inoculations, establishment of one variant was reduced. Mixed inoculations from two singly infected source plants resulted in fewer mixed infections than expected by chance. Therefore, the observed outcome was determined subsequent to host inoculation rather than in the vector. The outcome may be due to resource competition between pathogens. Alternatively apparent competition may be responsible; the pathogens' differential ability to overcome host defenses and colonize the host may determine the final outcome of new infections. Detailed knowledge of interactions between pathogens during transmission and establishment could improve understanding and management of disease spread.

Several mealybug species are vectors of grapevine leafroll-associated viruses (GLRaV), which cause the economically important grapevine leafroll disease in grape-producing regions worldwide. The mealybug Ferrisia gilli Gullan is a new pest of grapevines in El Dorado County, located in the Sierra Foothill wine-growing region of California. GLRaV species 1, 2, 3, and 4LV have been detected in vineyards with symptomatic vines in the Sierra Foothills. We conducted controlled virus acquisition and transmission experiments using source vine accessions infected with different combinations of GLRaV. We determined that F. gilli acquired GLRaV 1, 2, 3, and 4LV, and transmitted GLRaV-3 and GLRaV-4LV to uninfected recipient vines. Like numerous other mealybug species, in addition to causing direct damage to vines, F. gilli poses a threat to the grape industry as a vector of economically damaging viruses.