Secondary Logo

Journal Logo

Characterizing the Spatial Distribution of Ambient PAHs Using Vegetation Biomonitoring

Noth, E M; Hammond, S K; Biging, G S; Tager, I B

doi: 10.1097/01.ede.0000340525.60546.46
Abstracts: ISEE 20th Annual Conference, Pasadena, California, October 12–16, 2008: Contributed Abstracts
Free

University of California, Berkeley, Berkeley, CA, USA.

Abstracts published in Epidemiology have been reviewed by the organizations of Epidemiology. Affliate Societies at whose meetings the abstracts have been accepted for presentation. These abstracts have not undergone review by the Editorial Board of Epidemiology.

ISEE-1550

Back to Top | Article Outline

Background:

The use of Jeffrey pine needles as passive monitors to evaluate the spatial variability of polycyclic aromatic hydrocarbon (PAH) concentrations in ambient air in Fresno, California is an innovative approach to collecting large numbers (approximately 100) of samples from all areas of a city in less than one day to characterize spatial variability. Air sampling for PAHs is a labor-, equipment-, and time-intensive task which has yet to be achieved on a large scale. Using a biomonitor is a less intensive approach to characterizing the spatial variability. PAHs exist in ambient air as vapors (gas-phase) and adsorbed to particulate matter (particle-phase). Once emitted, they are deposited onto or absorbed into soil, vegetation, or surface water. Upon deposition to plants, PAHs partition into lipid-containing compartments of plants, such as leaves or needles. This partitioning behavior allows certain types of vegetation to be utilized as passive samplers for PAH, including pine needles. In order to quantify the PAH in pine needles, they are analyzed in a multi-step extraction and clean-up procedure. Samples are extracted in heptane, under 1500 PSI and at 120 degrees Celsius with an Accelerated Solvent Extractor. Then the extract is fractionated using a gel permeation column, in order to separate the organic matter from the PAHs. Last, the fraction is analyzed using GC/MS.

Back to Top | Article Outline

Methods:

Using a stratified random sampling approach, 158 pine tree locations were identified in Fresno. The sample size was reduced in two ways. First, trees closer than 10 meters to another tree were eliminated. Second, the traffic and housing characteristics were graded on a scale of density and proximity. The sample was reduced to 100 by eliminating and equalizing the numbers in each traffic and housing category. We attempted to collect samples at these 100 locations, but at 9 locations the tree was inaccessible due to recent tree trimming or locked gates. These data were used to design a spatial model of PAH distribution in Fresno, CA. This was done in two ways. First, a land use regression approach was taken. The dependent variable was the individual PAH values at each sampling location; the independent variables were topography, traffic density and proximity, and land use patterns. The second approach was to use kriging, a stochastic modeling approach, to interpolate values at non-sampled locations. These two outcomes are compared.

Back to Top | Article Outline

Results:

In the example of pyrene, the range of samples was from 6.4 ng/gram fresh needle weight to 20.4 ng/gram fresh needle weight. The average was 12.7 ng/gram fresh weight, and the standard deviation was 4.4 ng/gram fresh weight. Samples collected closer to the highways had higher levels, as well as samples collected along major arterials. Maps of both sampling and modeling outcomes will be presented.

© 2008 Lippincott Williams & Wilkins, Inc.