LB 13-121 - Foliar pathogen dilution due to species richness and phylogenetic dispersion increases productivity in diverse grassland communities

Biodiversity can benefit from pathogen dilution, as disease occurrence and adverse effects are expected to decline with compatible host density, but whether pathogen dilution occurs across systems is controversial. Plant foliar pathogens are particularly amenable to testing dilution and experimental manipulations have demonstrated dilution with increased diversity; however, it remains unclear how phylogenetic relatedness of available hosts effects this relationship. Additionally, functional consequences of foliar pathogen dilution remain underexplored. Despite evidence pathogens can mediate coexistence via plant-soil feedbacks, foliar pathogen dilution has not been linked to productivity benefits observed with diversity, i.e. overyielding. To test whether foliar pathogen dilution changes with host phylogenetic relatedness and contributes to overyielding, we established a grassland diversity manipulation experiment: altering richness (1, 2, 3, 5, and 6 plant species from 18 species) and phylogenetic relatedness (mixtures from one plant family, Asteraceae, Fabaceae, or Poaceae. or multiple families); replicated diversity combinations receive 50% or 150% ambient precipitation. Foliar disease incidence (DI) was assessed qualitatively (scaled 1-10). We hypothesized 1) foliar pathogen dilution would increase with plant richness, augmented in high precipitation, 2) disease incidence would increase in single-family mixtures, 3) foliar disease dilution would predict overyielding, specifically complementarity. DI decreased with plant richness in nine species (mixed linear model, p< 0.05), three Fabaceae, two Asteraceae, three Poaceae. 6sp had stronger dilution when phylogenetically over-dispersed compared to under-dispersed (e.g. D.canadida, p< 0.001). E.altissimum had stronger dilution in 50% ambient rainfall (p=0.05). To predict pathogen release, we calculated mean observed DI minus expected DI determined from monocultures. We estimated observed DI using linear regression slopes for each species’ DI response to species richness, adjusting for significant interactions with phylogenetic relatedness and precipitation. Pathogen release for each plot was the difference between average projected disease for each species from average monoculture DI for each species. Predicted pathogen release increased with complementarity (p < 0.001). Our findings contribute to evidence for foliar pathogen dilution, provide context dependence and consequences of dilution. Richness was enough to dilute disease for some species, but for others significant dilution only occurred in phylogenetically over-dispersed treatments. Dilution in response to precipitation was inconsistent across species, which could be determined by differing plant traits or tolerances for drought or high rainfall. The positive relationship between complementarity and predicted pathogen release suggests pathogens partially determine of biodiversity-productivity relationships. Our findings provide additional evidence of foliar pathogen dilution, that dilution is affected by phylogenetic relatedness and mediates overyielding.