8180

• UCREFRP Catalog Number: 8180
• Author: Hamilton, S. J.
• Title: Hypothesis of Historical Effects From Selenium on Endangered Fish in the Colorado River Basin.
• USFW Year: 1999.
• USFW - Doc Type: \
• Copyright Material: YES

Hypothesis of Historical Effects from Selenium

by selenium upon crop production." These high concentrations of selenium
in drain water are not unusual because Presser and Barnes (1985) reported up
to 1400 Ilg/L in drains emptying into the San Luis Drain, CA, in 1983; Crist
(1974) reported 1200 Ilg/L in drain water at the Kendrick Reclamation
Project, \iVY, in 1969; and Beath (1943) reported 1600 Ilg/L in a spring near
Thompson, UT. The criterion of the U.S. Environmental Protection Agency
for classifying selenium as a toxic waste is 1000 Ilg/L (USEPA, 1980); however,
wastes associated with agriculture are exempt (VSEPA, 1987b).
Historically, high selenium concentrations also w~re reported by Byers
(1936) in sediment samples. A sample from a sand and clay bar in the
Uncompahgre River 0.8 kIn above Delta, CO, contained 9 Ilg/g (assume wet
weight). A deposit from a drainage ditch 0.8 kIn north of Delta contained 1500
Ilg/g and a deposit from another drainage ditch 23.2 kIn east of Delta con-
tained 28,000 Ilg/g. Ooze in a main lateral ditch 0.6 km south of Chipeta
contained 4Ilg/g.
In 1936, additional water samples collected daily, but composited monthly
for analysis (April to September), from the Colorado River at Grandjunction,
CO, contained 0 to 10 Ilg/L (mean 4 Ilg/L), from the Gunnison River at
Grand junction contained 5 to 55 Ilg/L (mean 23 Ilg/L); whereas water
samples from the Colorado River at Cameo, CO, above irrigation inputs
contained 0 to 1 Ilg/L (Byers et at., 1938). Anderson et at. (1961) reviewed
these data and concluded that in the Gunnison River, the April sample
(15 Ilg/L) represented a spring "flushing-out of salts deposited in the fall on
the walls and floor of drainage ditches." The variation over time suggested to
Anderson et at. (1961) "an increasing return of seleniferous irrigation-drain-
age waters to the [Gunnison) river in july [25 Ilg/L), August [25 Ilg/L), and
September [55Ilg/L]," whereas high river runoff from spring melt waters kept
selenium concentrations low in May [5Ilg/L] andjune [10 Ilg/L]. During this
time period, Anderson et at. (1961) calculated that 73 to 178 kg/day of
selenium were discharged from the Gunnison River to the Colorado River in
1936. By comparison, Engberg (1999) calculated that the Gunnison River
passed 31 kg/day to the Colorado River in the period 1985 to 1994, which is
less than half of historic loading. Between the time of the computations by
Anderson et at. (1960) and the computations of Engberg (1999), significant
salinity control measures have been instituted by DOl and DOA in the Gunnison
River basin. It is speculated that salinity control has the added bonus of
reducing selenium loads (R. Engberg, personal communication, 1999). Nev-
ertheless, the 31 kg/day estimate was 1. 4 times and the 73 to 178 kg/ day
estimate was 3.4 to 8.2 times higher than the selenium loading (21. 7 kg/day)
that caused ecotoxic problems at Kesterson Resexvoir, CA (Presser and Piper,
1998) .
Historically, high selenium concentrations in water have been reported in
the southern part of the upper Colorado River basin. In 1936, headwaters of
the San Juan River had less than detectible concentrations of selenium, and
irrigation supply water in the Farmington and Shiprock, NM, area had sele-
nium concentrations of Illg/L (Byers et at., 1938). However, the Sanjuan

Hum. Ecol. Risk Assess. Vol. 5, No.6, 1999

1171