Rapid urbanization has profoundly impacted ecosystem patterns and processes in and outside of cities. These include changes in nitrogen (N) biogeochemistry, ecological community composition and wildlife behavior. Much less known, however, are the feedbacks between changes in behavior of wildlife and urban N cycling. Here we studied novel linkages between the roosting behavior of American crows (Corvus brachyrhynchos), crow-mediated input of N, the emergence of biogeochemical hot-spots and hot-moments, and the alteration of N biogeochemistry in urban ecosystems. We hypothesized that concentrated N input from crow roosts would change N cycling, in particular, shifting the balance between gaseous N2O and N2 outputs. Crow movement and roost density were monitored in the cities of Ithaca, Auburn, and Cortland in central New York from 2014-16. In April, July, October 2015 and March 2016, we collected intact soil cores from 8 woodland sites and 1 housing complex, and measured fluxes of N2O, N2 and CO2 in a gas-tight, free air recirculation system that allows for simultaneous quantification of all three gases by gas chromatography. Soil moisture, organic matter content and KCl extractable NH4+ and NO3-, were also quantified using a companion soil core collected at the same time.
Concentrated N input from crow roosts dramatically changed N biogeochemistry, raised soil NH4+ and NO3- concentrations and created hot-spots and hot-moments of N2O and N2 loss. At the Auburn crow (AC) site, where >40,000 crows roosted from December to March, fluxes of N2O and N2 in April 2015 averaged 60,222 and 53,410 µg N/m2/h, respectively, 1833 and 30 times higher than at a nearby reference site. Soil NH4+ and NO3- concentrations were 31-47 times those at the reference site. At the Ithaca and Cortland roost sites, where fewer crows were spread over larger areas, N2O fluxes were much lower (562.0 and 528.5 µg N/m2/h), although still 100 times higher than at non-roost sites. By July 2015, 4 months after the break-up of the winter crow roost, N2O and N2 fluxes at the AC site were still 40 and 4 times those at the reference site. Across the seasons and sites, N2:N2O ratio spanned a wide range, with a median of 47.1 and an inter-quartile range from 12.2 to 112.3. The N2:N2O ratio decreased exponentially with increasing soil N concentration, suggesting that enhanced wildlife-mediated N input increases N2O flux at the expense of N2 flux.