2006 Progress Report: Using Carbohydrates as Molecular Markers to Determine the Contribution of Agricultural Soil to Ambient Fine and Course PMEPA Grant Number: R832164
Title: Using Carbohydrates as Molecular Markers to Determine the Contribution of Agricultural Soil to Ambient Fine and Course PM
Investigators: Fraser, Matthew P.
Institution: Rice University
Current Institution: Arizona State University - Main Campus
EPA Project Officer: Chung, Serena
Project Period: December 1, 2004 through November 30, 2007 (Extended to November 30, 2010)
Project Period Covered by this Report: December 1, 2005 through November 30, 2006
Project Amount: $441,299
RFA: Source Apportionment of Particulate Matter (2004) RFA Text | Recipients Lists
Research Category: Particulate Matter , Air Quality and Air Toxics , Air
The overall goal of the proposed project is to fully develop, employ and verify a technique to quantify the contribution of agricultural soils entrained in the atmosphere to ambient fine and coarse particulate matter (PM). This project will test the hypothesis that the carbohydrate species present in agricultural soils are chemically distinct from organic components in native soils as a result of soil improvements designed to raise the organic content and productivity of agricultural soil. This project will focus on comparison of the concentrations of marker species in agricultural soils to unimproved soils, and attempt to separate agricultural emissions from other fugitive dust sources (such as windblown dusts, unpaved road dusts or construction dusts).
Year 2 research focused on conducting field measurements of fine particulate matter and local soil for analysis of carbohydrate concentrations. In addition, further work on extraction method comparison was conducted by a new graduate student assigned to the project, Yuling Jia.
For comparison of extraction methods, two methods were compared. The two extraction techniques focused on a well established method developed to sequentially recover non-polar and polar compounds (hexane followed by 2:1 benzene:isopropanol) as well as sequentially extracting with polar solvents (dichlormethane followed by methanol). For this study, ambient PM2.5 collected at two monitoring locations in Texas were extracted by each method. All sugars were quantified by GCMS as their trimethylsilylester derivatives.
Table 1. Comparison of Extraction Efficiency and Corresponding Sugar Concentrations in Aerosol Samples Using Two Extraction Methods.
Note: Method 1 extraction efficiency of San Augustine and Dallas samples were averaged to calculate the aerosol sugar concentrations of Clarksville samples using the same method.
Method 1: 2-30 ml aliquots of n-hexane and 3-30 ml of 2:1 benzene: isopropanol
Method 2: 3-15 ml aliquots of dichloromethane and 3-15 ml of methanol
Recovery: expressed at a significant level of α=0.05 (two-tailed)
RSD: Relative standard deviation
The extraction efficiency for the more polar method using methanol and dichloromethane resulted in improved extraction efficiency (between 76% and 81%) compared to the hexane plus benzene:isopropanol method (between 13% and 23%). Also note that while both extraction methods used spiked isotopically labeled glucose to monitor and to correct for extraction efficiency, the fact that extraction efficiency for the hexane and benzene:isopropanol method was low resulted in increased uncertainty for ambient sugar concentrations based on that method.
Particulate Matter Sampling and Analysis
High volume PM2.5 samplers were deployed to three sites in Texas to collect samples of fine particulate matter for quantification of sugar compounds. The sites included Dallas (urban), San Augustine (rural East Texas) and Clarksville (rural North Texas). Samples of fine PM were collected every third day from November 2005 through July 2006. This period allowed investigation of seasonal variations from the winter, through the spring growing season and into the summer. Also, it allowed the investigation of the impact of wildfires on ambient sugar concentrations in PM2.5 as the region was impacted by wildfires in the Texas Panhandle between November 2005 and March 2006.
Figure 1: Map of Monitoring Locations in Texas.
Levoglucosan was the most abundant sugar at all three sites (14-82% of the total sugars). Concentrations of levoglucosan were highest in winter and early spring, and it contributed 47- 82% to the total sugars during this period. This percentage dropped greatly to less than 30% when it entered into May and further decreased to less than 20% in July. Levoglucosan has been a reliable indicator of biomass combustion, and concentrations of levoglucosan in our samples were reflective of the local biomass burning, as the great enhancement of levoglucosan happened concurrently with the Texas-Oklahoma wildfires series.
Glucose was the second most abundant sugar (5% to 39%). As the growing season proceeded, concentrations of glucose in Clarksville increased from 10 ng/m3 in January to a maximum of 43 ng/m3 in May. After that, a strong decrease followed, to a much lower concentration of 12 ng/m3 in July. The same trend was observed for San Augustine and Dallas site, although less apparently. These changes seem to reflect the sugar production in ecosystem, with major synthesis of primary sugars early in the growing season (Medeiros et al., 2006).
Sucrose was significant only in the spring (March and April). It is the predominant sugar in the phloem of plants and is particularly important in developing flower buds (Bieleski, 1995). The strong decrease of sucrose amounts after April was probably due to the translocation of sugars to other parts of the plants for supporting growth in spring (van Doorn, 2004).
Trehalose and saccharide polyols have been documented as useful tracers for soil input to atmospheric particles in regions impacted by erosion or agricultural activities (Simoneit et al., 2004a). Little trehalose was detected in early spring, concentrations of trehalose significantly increased as it proceeded into summer season, demonstrating a good portion of aerosol trehalose coming from biogenic sources like soil and associated microbiota, since trehalose is an important fungal metabolite and has found to be only prevalent in early autumn soils (Simoneit et al., 2004a; Medeiros et al., 2006b). This trend was more established for San Augustine and Clarksville site than for Dallas site. This can be explained by an assumption that soil is a larger contribution factor to aerosol sugar contents at the rural site than at the urban site.
Figure 2. Monthly Individual Concentrations of Levoglucosan, Glucose, Sucrose and Trehalose in Aerosol Samples Collected at Three Sites in Texas (San Augustine, Clarksville and Dallas) from November 2005 to July 2006.
We are still working on the analysis of soil samples collected simultaneously with the fine PM at rural sites in Texas. This will allow a precise determination of the soil contribution to fine PM at these sites.
Subsequently, efforts will be on collecting fine PM, coarse PM and soil samples from a region impacted by agricultural emissions. This will likely be either California or Arizona and will occur in the fall of 2007.