springs but significantly lower in August 1993
The mechanisms by which WP particles enter
compared to August 1992. This decrease in mortal-
the pond sediments are unknown but could
include smoke projectile air bursts or ground
ity could be the result of lower WP availability, dif-
bursts, as well as leakage or subsurface explosion
ferences in sample size, human disturbance, or a
of duds. In the laboratory a burning particle of
combination of these factors.
WP dropped into water contained a significant
We found evidence of at least three, most likely
amount of unoxidized WP. Evidence is also pre-
four, eagle carcasses in the woods and on the Flats.
sented that WP may enter the sediments as a
Because this was the first systematic search of the
result of the decomposition of a WP-poisoned
woods surrounding ERF, the results suggest that
duck carcass.
WP contamination may have a greater impact on
A literature review of WP remediation tech-
bald eagles than previously thought, although
niques showed that ERF is the first documented
none of these carcasses had any analyzable tissues.
case of a U.S. artillery training area contaminated
Reitsma, L.R., and B.B. Steele (1995) Waterfowl
from wastewater, WP was oxidized by exposure
use and mortality at Eagle River Flats. In Inter-
to oxygen or chemical oxidants. In our laboratory,
agency expanded site investigation: Evaluation of white
air-drying and aeration of ERF sediments reduced
phosphorus contamination and potential treatability at
WP concentrations after several days. Oxidation
Eagle River Flats, Alaska (C.H. Racine and D. Cate,
Ed.). CRREL Contract Report to U.S. Army, Alaska,
tion. In ERF a test explosion simulating the deton-
Directorate of Public Works, FY94 Final Report, p.
ation of a 105 howitzer projectile was shown to
289320.
redistribute, not oxidize, WP particles; thus,
The main objective of the waterfowl mortality
resumption of summer firing is not a remediation
work in 1994 was to continue measuring mortality
option. A method of monitoring waterfowl mor-
with the standardized method established in 1991
tality in permanent transects was developed and
and further revised in 1992. In addition, habitat
tested and should be used to establish a baseline
use studies were conducted for input into the
mortality index for evaluating the success of
IRAM (Integrated Risk Assessment Model)
future remediation.
described below. Photo-identifications of scaven-
ger species that remove WP-poisoned carcasses
Reitsma, L.R., and B.B. Steele (1994) Waterfowl
from ERF were also conducted.
use and mortality at Eagle River Flats. In Inter-
Carcasses were counted on transects in Areas A,
agency expanded site investigation: Evaluation of
C, the Bread Truck Pond and Racine Island during
white phosphorus contamination and potential treat-
the spring and fall migration periods. We also
ability at Eagle River Flats, Alaska (C.H. Racine and
counted feather piles on transects in the bordering
D. Cate, Ed.). CRREL Contract Report to U.S.
woods in the spring. These feather piles represent
Army Garrison, Alaska, Directorate of Public
carcasses removed from the Flats by scavengers
Works, FY93 Final Report, p. 205226.
(mostly eagles). Mortality was calibrated by daily
The objective of the waterfowl mortality study
censuses to quantify use of ERF by migrating
was to assess the relative amount of waterfowl
waterfowl over the same period. These data were
used in an ANCOVA (Analysis of Covariance) to
mortality in order to detect year-to-year changes
as WP exposure decreases through remediation
analyze for significant differences in mortality
efforts. Our approach over the past 3 years (1991
between years. The mortality rate in the spring of
1993) has been to count feather piles and car-
1994 was significantly lower than in the springs of
casses on sample areas of ERF and in the sur-
1993 or 1992. The mortality rate in fall 1994
rounding woods and to extend the total area sam-
remained lower (as in fall 1993) than in fall 1992.
pled to more accurately estimate total mortality
Reasons for these differences are discussed. Water-
and detect potential new areas of high mortality.
fowl habitat use was quantified by conducting
We found that overall mortality has varied
simultaneous observations of waterfowl behavior
annually. We now have complete and comparable
from two to four locations in spring and fall. These
data for 1992 and 1993, and we found that mortal-
observational data can be used to assess habitat
ity was much lower in 1993 than in 1992. This
preferences and to calculate risk using the Inte-
lower mortality is mostly a reflection of the fact
grated Risk Assessment Model described below.
that fewer waterfowl used ERF in 1993. However,
Preliminary photographs at carcass-baited trip
the mortality was proportionally similar in both
mechanisms wired to cameras in the woodland
60
to contents
Previous Page
