 |
||
|
|
||
| |||||||||||||||
|
|
 ERDC TN-DOER-R5
September 2004
methods are explained in the references cited, and results obtained with each method are
reported.1 The RSS computational method is relatively simple and easily incorporated into a
spreadsheet. This method combines systematic method error2 and propagated measurement error
(random and systematic) to calculate total error (TE). The range of uncertainty is the value of
TBP " TE. For PAHs, TBP uncertainty by RSS was less than a factor of two in most cases, and
was always less than a factor of five. The bootstrap method is computer intensive but is more
precise, and provides nonparametric statistical uncertainty measures such as standard errors and
confidence intervals. Tests of significance can also be performed. Using bootstrap statistics on
the same PAH data set, TBP uncertainty was always less than a factor of two. These methods are
similarly applied to calculate TBP uncertainty for PCDD/Fs in the present study.
METHODS:
Experimental Design. Dioxin and dibenzofuran concentration data of coastal shelf sediments,
organic carbon content of the sediments, lipid content of the biota collected with the sediments,
and BSAFs calculated from data contained in the USACE ERDC BSAF database3 provided input
parameters for the TBP model:
TBP = BSAF*(Cs/fOC)*fL
(1)
where
Cs = concentration in sediments, pg g-1, dry weight
fOC = decimal fraction sediment organic carbon, dry weight
fL = decimal fraction organism lipid, wet weight
BSAF = biota/sediment accumulation factor (unitless)
TBP = theoretical bioaccumulation potential, pg g-1, wet weight
Uncertainty was computed for TBP using both the RSS and bootstrap methods. Tissue
concentrations of PCDD/Fs measured in the biota collected with the sediments were then
compared with the predicted concentrations (TBP) and their ranges of uncertainty.
Field Collections. Sediments and resident organisms were collected using a 0.1-m2 Smith-
McIntyre grab sampler in the New York Bight Apex, at an area surrounding 40E20.48' North,
73E52.34' West during 20-24 August, 1991. Sediment samples were emptied into plastic sorting
trays prepared by acid washing, hexane rinsing, and rinsing with site water prior to use. A 0.5-L
aliquot of sediment was removed from each grab sample and composited. At the end of each
sampling day, the composited sediments were homogenized, large organisms were removed by
hand, and five 1-L replicate samples were placed in acid-cleaned, hexane-rinsed glass jars with
teflon lids. Care was taken to ensure that no head space remained in the containers. The samples
were stored under refrigeration (4EC) until the end of the field sampling effort, then shipped in
1
McFarland and Clarke (1999) is available in pdf format on the USACE Engineer Research and Development
Center website: http://www.wes.army.mil/el/dots/eedptn.html . Readers are referred here for explanations of
TBP, BSAFs, and calculation of RSS uncertainty.
2
Systematic method error is the error inherent in the model itself, were all input data measured with perfect
accuracy. It is estimated by comparison of model predictions to actual bioaccumulation.
3
The Engineer Research and Development Center Biota/Sediment Accumulation Factor (BSAF) and Lipid Database
can be found at: http://www.wes.army.mil/el/dots/database.html
2
|
|
Privacy Statement - Press Release - Copyright Information. - Contact Us - Support Integrated Publishing |
