Urban soil heterogeneity and plant-soil interactions in New York City parkland

Background/Question/Methods

Spatially variable social-ecological systems in urban environment such as New York City (NYC) create complex heterogeneous soil and ecosystems.  In the NYC region, patches of native, manmade, and degraded ecosystems coexist, resulting in more variable vertebrate and fungal densities (Pouyat et al., 1994), soil chemistry (Pouyat et al., 1995), and soil trace gas fluxes (Goldman et al., 1995) than ecosystems in rural areas.  In collaboration with NYC Parks and Recreation, we have established experimental reforestation plots around NYC to investigate 1) spatial variability of soil parent material, nutrient availability, carbon sequestration, and toxicity; 2) urban plant-soil interaction; and 3) the influences of management practices, diversity, and initial soil characteristics on the outcome of reforestation efforts in urban parkland.  We established subplots with high and low diversity tree plantings, with or without shrub plantings at eleven locations with varying land use history, vegetation cover, diversity, and hydrology. We monitor plant diversity, invasive species expansion, and plant-soil interactions in 38 subplots.  Thirty-nine soil chemical parameters, including nutrient availability (i.e., N and P), toxic metals (i.e., Pb and Zn), lithogenic indicators (i.e., Ti and Ca), carbon, and pH were qauntified over three-year period.  Multivariate analyses were used to identify patterns of soil variation.

Results/Conclusions

Soils around NYC are found to be highly heterogeneous.  Cluster analyses suggest NYC Park soils may be separated into at least 4 distinct groups based on their chemical composition that differ in lithology, trace metal concentration, carbon content, and nutrient availability.  High variability of lithogenic composition (Fe, Ti, Ca, and K) suggests that most soils were brought in from various sources.  Regardless of spatially heterogeneous soil matrix and toxic metal concentrations, Pb, Zn, Cu, and Cr greatly covaried from site to site, explaining 41% of the variance of trace metal distribution, suggesting a common pollution source at most sites.  However, variable Pb/Zn ratios suggest non-homogeneous pollution sources at some sites. Nutrient parameters (N, P, pH and Ca) paralleled one another and may have influenced vegetation composition. Within one year, 90% of the sites have increased nitrate or phosphorus with an average increase of 27%. Soil carbon ranges from 1.7 to 7.4%, with the highest values found in sites with the densest vegetation.  After one year, 85% of subplots exhibit an increase of soil carbon (40% average) with the highest carbon increase at the site with the highest native species coverage and diversity.  The results provide framework for urban ecosystem restoration.