Dartmouth College Chicago, New Hampshire, United States
Natural soils can take thousands of years to form and house distinct communities of plants and soil microbes depending on the physical properties and environmental conditions of the soil. In contrast, urban soils are often newly engineered, reclaimed, or highly degraded from anthropogenic activities. These soils often have disparate nutrient loads and altered hydrology making native plant establishment challenging. Arbuscular mycorrhizal (AM) fungi may aid in urban plant reestablishment as they deliver scarce soil resources and other physiological benefits to plants in exchange for photosynthetically-derived carbon. However, little is known about how to restore AM fungi to urban soils, how communities form without amendment, and how long it takes for urban AM fungal communities to establish. In this study, we investigate active and passive mechanisms of establishing urban AM fungal communities in experimental green roofs. We used next generation sequencing to determine AMF community shifts over time compared to species pools present in local native grassland inoculum and aerial samples taken from the roof.
Across the study, amplicon sequence variants corresponding to 63 virtual taxa were detected. Of these taxa, 29 were found in inoculated plots and 24 were present in uninoculated plots over the sampling period. Passive reestablishment of AM fungi onto green roof soils via aerial dispersal likely occurred for a single taxa, a weedy species, Paraglomus Pa 1 found in both aerial and green roof samples. Paraglomus species typically have small diameter spores, supporting evidence that AM spore traits predict aerial dispersal capabilities. Eleven species were found in the inoculum and control plots, but not the aerial samples, suggesting short distance dispersal from the inoculated plots into the controls. As expected, a high percentage of the taxa detected in the inoculum were also found in the inoculated plots, increasing from 65% to 70% two years after the initial inoculation date, suggesting that the establishment of some inoculum species takes time. We propose that multiple vectors of inoculation of engineered urban soils are possible, but that inoculation can have a strong, immediate effect on the fungal community. Stochastic processes like dispersal affect local and regional species pools that are important to consider in the recovery of AM fungal communities in degraded soils.