Central Mineral and Environmental Resources Science Center
Subtask Contact: David B. Smith
At the 2003 Soil Geochemistry Workshop, it was recognized that pilot studies were necessary to test and refine the recommended sampling and analytical protocols for the proposed soil geochemical survey of North America. We chose to conduct these studies at both a continental scale and a regional scale. In 2004, we initiated work on a regional-scale pilot study area in northern California and on a continental-scale study of two transects across the United States and Canada. The purpose of this section was to 1) outline the rationale for continental-scale pilot studies, 2) describe the sampling and analytical protocols used, and 3) present some preliminary results.
Two continental-scale transects were chosen for study as shown in Figure 1 below.
Figure 1. Map showing soil sample sites along the two continental transects. [Large version of Figure 1]
One transect extended from northern Manitoba and currently ends near El Paso, Texas. The Servicio Geológico Mexicano extended this transect into Mexico during 2005-2006. The other continental-scale transect approximately follows the 38th parallel and extended from just north of San Francisco to the Maryland shore. The transects crossed multiple geologic, climatic, physiographic, land use, pedologic, and ecological boundaries. This imposed rigorous field testing of sampling protocols across a broad range of conditions.
There were several advantages to conducting pilot studies along such transects. 1) The N-S transect gave a clear indication of international collaboration. 2) There was a larger knowledge gap concerning microbiological and organic compounds than trace element geochemistry. If we did not delineate a gradient along such continental-scale transects, the merit of the project to map microbiological and organic parameters would have been in doubt. We needed to know this sooner rather than later. 3) Conducting pilot studies along such transects enabled us to test the sampling and analytical protocols and to optimize field logistics from permafrost of northern Canada through desert environments in the southwestern US, thus allowing for a smoother and more efficient path forward to the soil geochemical survey of all of North America. In addition, the pilot was designed so the sampling could be completed in one field season (2004) and chemical analyses and microbiological characterization could be completed by spring of 2005. This accelerated schedule allowed us to begin making presentations and producing publications in summer of 2005.
The global Geochemical Reference Network (GRN) was used as a basis for selecting sample sites along the transects. The GRN is a network of approximately 5,000 grid cells (about 160 x 160 km in size) distributed over the land surface of the Earth and was established to provide a basis for conducting a global-scale geochemical survey (Darnley and others, 1995). Figure 2 shows the GRN for portions of Canada, Mexico, and the United States.
Figure 2. Map showing the global geochemical reference network (GRN) for parts of Canada, Mexico, and the United States. [Large version of Figure 2]
To select sample sites, each transect was divided into approximately 40 km segments. A 1-km-wide strip was randomly selected for each segment; within each strip, a site reflecting the most representative landscape within the dominant soil type was chosen for sampling. Duplicate samples were collected 10 meters apart to estimate local spatial variability at one in four sites. The N-S transect consists of 105 sites and the E-W transect has 160 sites.
The focus of these transects was the anthropogenic component of the geochemistry versus the natural (geologic) component. At each site, a modification of the recommended sampling protocols from the Soil Geochemistry Workshop was used. Detailed sampling protocols [PDF file, 48 KB] and a field form [PDF file, 32 KB] were developed prior to the beginning of field work. Up to four samples were collected at each site: 1) surface soil from 0-5 cm depth; 2) O-horizon, if present; 3) a composite of the A-horizon; and 4) a composite of the C-horizon. The first three sample types provide an indication of anthropogenic input while the C-horizon sample provides information on the natural (geologic) input.
Each sample collected was air dried, sieved to minus 2 mm, ground, and analyzed for major and trace-element composition by the following methods: (1) inductively coupled plasma-mass spectrometry (ICP-MS) and inductively coupled plasma-atomic emission spectrometry (ICP-AES) following a near-total four-acid digestion for aluminum, antimony, arsenic, barium, beryllium, bismuth, cadmium, calcium, cerium, cesium, chromium, cobalt, copper, gallium, indium, iron, lanthanum, lead, lithium, magnesium, manganese, molybdenum, nickel, niobium, phosphorus, potassium, rubidium, scandium, silver, sodium, strontium, sulfur, tellurium, thallium, thorium, tin, titanium, tungsten, uranium, vanadium, yttrium, and zinc; (2) cold vapor- atomic absorption spectrometry for mercury; (3) hydride generation-atomic absorption spectrometry for antimony and selenium; (4) coulometric titration for carbonate carbon; and (5) combustion for total carbon and total sulfur. An estimate of bioaccessibility was made by a deionized water leach on A-horizon soils and an extraction by a simulated human gastric fluid on 0-5-cm material.
A separate composite of the A-horizon was collected at each site for microbial characterization by phospholipid fatty acid analysis, BIOLOG analysis, enzyme assays, and agricultural and human pathogen screening. A separate sample of the 0-5-cm material was collected for determination of organic compounds such as pesticides, PAH's, and their breakdown products.
Sampling along these transects was initiated in June 2004 and completed in October 2004. A total of 265 sites were occupied and 1,558 samples were collected. The near-total major- and trace-element data were published in July 2005 as USGS Open-File Report 2005-1253 (http://pubs.usgs.gov/of/2005/1253/).
Plots showing concentrations of elements along the transects clearly demonstrate continental-scale geochemical patterns that are related to a combination of natural and anthropogenic processes. The natural processes controlling the distribution of elements in soil include formation of soils from weathering of parent material (bedrock) of varying composition and climate-driven processes that establish soil moisture regimes and levels of organic matter in soil. Anthropogenic influences such as industrialization, urbanization, waste disposal, mining, and agriculture regularly are superimposed on these natural, or background, geochemical distributions.
Figure 3 shows an example of an elemental distribution greatly influenced by soil parent material.
Figure 3. Distribution of chromium in C-horizon soils of both the N-S and E-W transect. The height of the bars (red for E-W and blue for N-S transect) is proportional to the chromium concentration (parts per million, ppm). [Large version of Figure 3]
The chromium content of soil in the Coast Ranges and foothills of the Sierra Nevada Mountains in California is highly elevated compared to the other samples collected along the E-W and N-S transects. This area of California is underlain, in part, by ultramafic rocks. These types of rocks contain higher concentrations of chromium, nickel, and cobalt in comparison with other rock types. Figures 4 and 5 show nickel and cobalt along the E-W transect in a different style of plot.
Figure 4. Cobalt concentration (parts per million) in C-horizon soils of the E-W transect.
Figure 5. Nickel concentration (parts per million) in C-horizon soils of the E-W transect.
Other examples of the influence of parent material can be seen in data from the N-S transect. A plot of manganese in both A- and C-horizon soils shows elevated values associated with the lacustrine sediments of glacial Lake Agassiz (Fig. 6.)
Figure 6. Plot of manganese concentration in A- and C-horizon soils along the N-S transect. The blue symbols (x) represent C-horizon samples and the red symbols (+) represent A-horizon samples. The solid blue line on the graph represents a statistically smoothed plot of the C-horizon data and the solid red line represents a similar plot for the A-horizon samples. The bold red line represents the approximate location of the N-S transect. [Large version of Figure 6]
Lead distribution in both A- and C-horizon soils along the E-W transect shows elevated values in southeastern Missouri and southern Illinois. We hypothesize that this lead enrichment is partially due to natural weathering of lead-rich rocks of the southeast Missouri lead district and partially due to 1) disturbance of these materials by mining activities and 2) emissions from lead smelters in the area.
Figure 7. Plot of lead concentration in A- and C-horizon soils of the E-W transect. The blue and red lines on the graph represent statistically smoothed plots of the C- and A-horizon data, respectively. The bold red line across the top of the chart represents the approximate trace of the E-W transect. [Large version of Figure 7]
Many elements bind strongly to organic material and are thus enriched in more organic soils. A plot of organic carbon in A-horizon soils along the E-W transect clearly shows a continental-scale pattern of higher concentrations east of central Kansas (Fig. 8).
Figure 8. Plot of organic carbon content (weight percent) of A-horizon soils along the E-W transect.
This pattern is likely related to the greater precipitation received in the eastern portion of the transect as compared to the western portion, which leads to more vegetative cover. We see a similar pattern for mercury in A-horizon soils (Fig. 9). Mercury is one of the elements known to bind with organic matter in soil.
Figure 9. Mercury concentration in A-horizon soils of the E-W transect.
The distribution of some of the major elements may also be controlled to some extent by climate. For example, calcium distribution shows a clear pattern of higher concentrations in the western part of the E-W transect (Fig. 10). We hypothesize that this element has been leached from the soils of the eastern part of the transect. This leaching is caused by a combination of lower pH of the soils (higher acidity) and the higher rainfall in the east.
Figure 10. Calcium concentration in A-horizon soils along the E-W transect.
Soil enzymes are the mediators and catalysts of important soil functions that include: 1) decomposition of organic matter and 2) release of inorganic nutrients for plant growth (Dick, 1997). These enzymes are often closely related to soil quality parameters such as biomass or microbial activity in addition to organic matter. Professor Richard Dick of Ohio State University analyzed A-horizon soils from both the E-W and N-S transects for two enzymes: arylsulfatase and Β-glucosidase. Arylsulfatase plays an important role in the release of sulfate, the plant available form of sulfur, from various organic sulfate esters (Tabatabai and Bremmer, 1970). Since microbial ester sulfates are only found in fungi and not in bacteria, elevated arylsulfatase may be related to fungal biomass (Dick and others, 1996). Arylsulfatase activity, expressed as mg p-nitrophenol kg-1 soil hour-1, clearly shows higher values in the eastern part of the E-W transect (Fig. 11) and in the northern part of the N-W transect (Fig. 12). We hypothesize that these patterns are related to the larger fungal biomass in those parts of Canada and the U.S. having more vegetative cover.
Figure 11. Arylsulfatase activity (mg p-nitrophenol kg-1 soil hour-1) in A-horizon soils of the E-W transect [Large version of Figure 11]
Figure 12. Arylsulfatase activity (mg p-nitrophenol kg-1 soil hour-1) in A-horizon soils of the N-S transect.
Estimates of bioaccessibility were made on A-horizon soils by a deionized water leach and by an extraction with a simulated human gastric fluid. The simulated gastric fluid consisted of a solution of deionized water, hydrochloric acid, and glycine (pH = 1.5). The percent of the total amount of an element in soil removed by the water leach varied from more than 10% to less than 0.001%. Selenium was the most bioaccessible element in the water leach, and titanium was the least bioaccessible (Fig. 13). This variability is a function of the solubility of the mineral phase in which each element resides within the soil.
Figure 13. Leachability of selected elements, expressed as Tukey boxplots, in a deioinized water extraction of A-horizon soils from both the E-W and N-S transects. [Large version of Figure 13]
The total amount of element extracted increased considerably with the simulated gastric fluid leach (Fig. 14).
Figure 14. Arsenic extracted by a deionized water leach (burgundy bars) compared to arsenic extracted by a simulated gastric fluid leach (gray bars). [Large version of Figure 14]
Seventy-three samples of 0-5-cm material from the U.S. portion of the N-S transect were analyzed for 19 organochlorine pesticides and breakdown products. Only three of the 73 samples analyzed had detectable quantities. Dieldrin was found in one sample from an agricultural field in Kansas (6.6 µg/kg) and one sample from an agricultural field in Nebraska (16 µg/kg). DDT (47 µg/kg) and DDE (50 µg/kg) were found in soil taken from a salt flat in southern New Mexico (Fig. 15).
Figure 15. Map showing location of three samples containing detectable organochlorine pesticide.
Darnley, A.G., Björklund, A., Bølviken, B, Gustavsson, N., Koval, P.V., Plant, J.A., Steenfelt, A., Tauchid, M., Xie, Xuejing, with contributions by Garrett, R.G., and Hall, G.E.M., 1995, A global geochemical database for environmental and resource management: Recommendations for international geochemical mapping, Final report of IGCP Project 259: Earth Science Series, v. 19, Paris, United Nations Educational, Scientific, and Cultural Organization, 122 p.
Dick, R.P., 1997, Soil enzyme activities as integrative indicators of soil health, in Pankhurst, C.E., Boube, B.M., and Gupta, V.V.S.R., eds., Biological Indicators of Soil Health, Oxford, Oxford University Press, pp. 121-156.
Dick, R.P., Breakwell, D.P., and Turco, R.F., 1996, Soil enzyme activities and biodiversity measurement as integrative microbiological indicators, in Soil Science Society of America, Methods for Assessing Soil Quality, SSSA Special Publication 49, pp. 247-270.
Tabatabai, M.A., and Bremner, J.M., 1970, Factors affecting soil arylsulfatase activity: Soil Science Society of America Proceedings, v. 34, pp. 427-429.