We conducted an observational study at 62 golf courses at varying distances from urbanized areas in the temperate, mesic U.S. mid-Atlantic region, measuring soil carbon stocks in minimally managed areas where cool-season turfgrasses had grown without disturbance for at least 25 years. In 2009-2010, soils were sampled to 30 cm, and site and management variables were recorded. Total and permanganate oxidizable soil carbon were quantified, and potential explanatory factors were explored using multiple regression analysis. The strongest interaction that can be seen in these observations is a positive, log-log, 'broken stick' or threshold relationship between soil Pb and total soil C.  Our analysis suggests that above a threshold of ca. 4 mg extractable Pb/kg soil, lead (Pb) was strongly and positively correlated with soil carbon.  Increasing minimum daily temperature in February was also positively correlated with soil carbon, and several other correlations can be observed. Large quantities of Pb have been distributed atmospherically over the last century. If the Pb effects observed here are widespread (e.g., in other regions and ecosystems), they could impact global carbon cycle modeling and/or uncertainty. A better understanding of Pb effects on soil carbon turnover, erodibility, and mineralization dynamics, particularly under disturbed or agricultural soils, is urgently needed, and we encourage other soil scientists to investigate.

Context of data collection:

We utilized undisturbed soils along an urban-rural transect throughout Maryland and Washington, D.C., in the mid-Atlantic region of the U.S. to study the complex interactions of multiple global change factors *in toto. *Fifty-two public and private golf facilities in continuous operation for > 25 years were identified with the assistance of the Mid-Atlantic Association of Golf Course Superintendents in early 2009. These businesses were generous in allowing us to collect soil samples and make other observations.  The latitude and longitude of each sampling location is included in these data, but the names of the golf courses are not.

Of the 52 golf courses located, seven facilities included two entire, separate courses constructed in different years, each of which was treated as a separate, independent sampling site. Three additional facilities included single courses that spanned two clearly different soil parent materials, ascertained by comparing auger samples (1.0 m depth) with area soil maps.  For these facilities, each of the two contrasting parent material areas were separately sampled and considered as independent sampling sites.  The total number of independent samples for the study was therefore 62.  Missing data are denoted with “NA”.

For purposes of soil carbon, we detected and excluded one outlier, which is the sample identified as "JW6".  Total soil carbon is significantly higher at this location, and the soil contained visible charcoal.  The manager confirmed verbally that debris from the 1904 "great Baltimore fire" was buried at this facility.   If you are reanalyzing this dataset, please consider this aspect of site JW6.

Brief Methods: 

·       Intact-core soil samples composited into 0-5, 5-15, and 15 – 30 cm depth increments

·       Bulk density, course fragment content, soil total C content and POX C content were measured separately in each of the three composited depth increments and weighted averages over the full 30 cm profile calculated mathematically when indicated; total C and POX C also reported in units of Megagrams C / hectare for 0 to 5 and 0 to 30 cm depths.

·       Soil chemical analyses were conducted by the University of Delaware Soil Testing Lab on composite samples representative of the entire 0 – 30 cm depth (i.e. one sixth of the sample taken from 0 to 5 cm, one third from 5 to 15 cm, and one half from 15 to 30 cm depth).  Therefore, there is only a single value per site representing the full 30 cm profile for these measurements. 

·       Permanganate-oxidizable C

·       Mehlich-3 soil extraction

·       Long-term soil temperature measurements via buried HOBO data loggers

·       Passive monitoring of ozone exposure via oxidation of indigo carmine dye

·       Exploratory multiple regression