Submitted to: Journal of Soil Contamination
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/1/1998
Publication Date: N/A
Interpretive Summary: Elevated levels of selenium have been measured in sediment at Kesterson Reservoir and contributed to the death and deformities in waterfowl frequenting this site. Strategies are not readily available for effective reduction of high Se levels. As part of a remediation strategy to lower levels of soil Se, canola was planted in Se-contaminated sediment collected from the Reservoir, as well as planted on-site. Canola lowered Se concentrations more effectively under greenhouse conditions than under field conditions. The study clearly shows that extrapolating results from greenhouse to field conditions should be done with caution. Using canola to lower Se levels at Kesterson Reservoir will be a long-term management strategy under field conditions.
Technical Abstract: Selenium (Se) may be present in soils and sediments in high concentrations, yet not be mobile or available for plant uptake. Phytoremediation of Se by canola (Brassica napus) was evaluated in sediment from Kesterson Reservoir collected from three different depths (0-30, 30-60, 60-90 cm) under greenhouse and field conditions. In the greenhouse study, total soil Se concentration at preplant ranged from 10 to 112 mg/kg DM. Shoot Se concentrations of canola were 182, 53, and 19 mg/kg DM in the 0-30, 30-60, and 60-90 cm depths, respectively. Percentages of Se accumulated by canola relative to total Se loss from the soil at postharvest were as high as 24%. In the field study, total soil Se concentrations were as high as 26 mg/kg DM soil. Field-grown canola accumulated approximately 50 mg/kg DM which accounted for less than 10% of total Se lost in the soil at postharvest. Phytoremediation of Se-laden soils under field conditions was about 50% of that observed under controlled greenhouse conditions. This relationship may be useful for prediction of field remediation operation using greenhouse generated data.