Together, results from field experiments and a literature review support the idea that increased pathogen loads will be a mechanism by which global change alters terrestrial ecosystem processes. All four experiments were conducted at Cedar Creek Natural History Area, MN.

Several lines of evidence suggest that the severity of many plant diseases is likely to increase with climate warming and be affected by changes in moisture. First, while suitable long-term data sets are uncommon, interannual variation in the severity of several crop and forest diseases is positively correlated with temperature and moisture, suggesting that they will also track directional climate change. Second, controlled experiments have documented that the growth and reproduction of numerous pathogens requires highly humid conditions occurring primarily overnight, and that growth and reproduction are often maximized at temperatures greater than current overnight lows. Third, outside the tropics, overwintering survivorship of many pathogens is near zero, causing a drastic population bottleneck that climate warming may alleviate. Finally, the most severe and least predictable disease outbreaks may occur if shifts in species' geographic ranges under climate change allow pathogens to infect novel hosts from which they were previously geographically isolated and which lack resistance.

In a factorial grassland experiment, three components of global change - elevated CO2, nitrogen addition, and decreased plant diversity - all increased pathogen load (percent leaf area infected by fungi) for much to all of the plant community. Elevated CO2 increased pathogen load of C3 grasses, probably by decreasing water stress, increasing leaf longevity, and increasing photosynthetic rate. Decreased plant diversity further magnified the increase in C3 grass pathogen load under elevated CO2. Nitrogen addition increased pathogen load of C4 grasses by increasing foliar nitrogen concentration. Decreased plant diversity had the broadest effect, increasing pathogen load across the plant community by allowing remaining plant species to increase in abundance, facilitating spread of pathogens specific to each plant species. Plant species composition also influenced community pathogen load: communities that lost less disease prone plant species increased more in pathogen load. Similar effects of diversity and composition were found the year before in a separate experiment in which diversity was the only manipulated variable. To test the effects of human alteration of fire regimes on disease and other processes, oak savanna plots averaging 15 hectares were experimentally burned at a range of fire frequencies for 36 years. The severity of nine foliar fungal plant diseases was generally reduced in the short-term by burning, presumably because of increased pathogen mortality. In contrast in the long-term, severity was generally increased by more frequent burning because burning increased abundance of herbaceous hosts and created a microclimate more favorable to disease.

To test the effects of changes in pathogen load on ecosystem processes, foliar fungal pathogens were experimentally excluded from grassland plots using fungicide. As predicted theoretically, pathogens strongly limited belowground carbon allocation by plants. Pathogen exclusion increased root production by 1/3, root biomass by half, and soil respiration by 18%. Pathogen exclusion did not increase aboveground production or biomass, but did increase community leaf longevity by 23% and net photosynthesis of the dominant species, Andropogon gerardii, by 32%. At peak, pathogens infected less than 9% of community leaf area. These results are the first to identify pathogens as regulators of ecosystem processes outside of crops and forests.

Together, these results suggest that global change will increase grassland pathogen loads, which in turn will decrease root production and biomass. Therefore, pathogens may help limit the ability of ecosystems to sequester carbon in a changing environment. cem46@cornell.edu.