Lack of trade-off between direct and indirect defence against grape powdery mildew in riverbank grape

1. The hypothesis that plants that invest more in an indirect defence will invest less in direct mechanisms was tested using genotypes of the riverbank grape Vitis riparia that varied in expression of acarodomatia, tufts of non-glandular trichomes on underside of leaves that are frequently occupied by mycophagous mites. Mycophagous mites in this system have been shown to protect vines from grape powdery mildew caused by the fungus Uncinula necator. 2. Multiple cane cuttings from 24 genotypes of V. riparia, a species native to North America, were obtained from the USDA grape genetics repository and planted out in a common garden. 3. Genotypes varied in size of acarodomatia by threefold and varied in density of mycophagous mites by over 10-fold. Genotypes originating from the western part of the distribution had smaller acarodomatia and fewer mycophagous mites than genotypes from the eastern part of the distribution. There was a positive genetic correlation between acarodomatia size and mite density. 4. In a separate experiment, where vines were kept free of mycophagous mites, genotypes showed 10-fold variation in percentage leaf surface colonised by mildew. There was, however, no relationship between mildew on leaves and size of acarodomatia. Mildew did not vary depending on origin of genotype, although female vines tended to have greater levels than males. 5. No evidence was found for a trade-off between investment in an indirect defence mediated through acarodomatia and investment in direct defence against grape powdery mildew. Mite/plant defence mutualisms are contrasted with ant/plant defence mutualisms where there is better evidence for trade-offs under field conditions.