TEMPORAL AND SPATIAL MONITORING OF SOIL SELENIUM AT KESTERSON RESERVOIR, CALIFORNIA

Citation
C. Wahl et al., TEMPORAL AND SPATIAL MONITORING OF SOIL SELENIUM AT KESTERSON RESERVOIR, CALIFORNIA, Water, air and soil pollution, 74(3-4), 1994, pp. 345-361
Citations number
27
Categorie Soggetti
Environmental Sciences","Water Resources
ISSN journal
00496979
Volume
74
Issue
3-4
Year of publication
1994
Pages
345 - 361
Database
ISI
SICI code
0049-6979(1994)74:3-4<345:TASMOS>2.0.ZU;2-P
Abstract
The selenium and salt content of the top 15 cm of the soil profile at Kesterson Reservoir (Merced County, California) have been monitored an nually to develop a data set that provides a foundation for: (1) evalu ating the status of the selenium inventory and biological hazards at K esterson Reservoir; (2) understanding selenium fluxes near the soil su rface; and (3) making long-term predictions of the selenium concentrat ions available for plant uptake and dissolution into rainwater ephemer al pools. Results of this monitoring program indicate that the soil se lenium inventory at Kesterson may be categorized in terms of three mai n patterns: (1) spatial trends associated with historic Reservoir oper ations; (2) temporal trends due to the oxidation and remobilization of the selenium inventory and; (3) temporal trends due to seasonal cycle s. It is evident that the selenium inventory and distribution within t he soil profile will evolve slowly whereby the fraction of the total i nventory that is now immobile (est. at 93%) will oxidize to more mobil e and bioavailable forms. Two major issues with broad importance were raised concerning sampling the surface soil selenium environment at Ke sterson which may be helpful to others conducting investigations of si milar nature. These issues include: (1) the recognition that variation s in surface soil contaminant concentrations due to seasonal redistrib ution may obscure long term trends and; (2) large spatial variability in soil contaminant concentrations make it difficult to obtain large e nough data sets to detect statistically significant changes in the con taminant inventory until large changes have already taken place. A com bination of both process-oriented and synoptic type sampling are recom mended to better define time trends.