Northern Prairie Wildlife Research Center
Effects of the Herbicide Atrazine on Ambystoma tigrinum Metamorphosis:
Duration, Larval Growth, and Hormonal Response
Material and Methods
Culture
Eyed-eggs of tiger salamanders were collected from a semipermanent wetland in
Stutsman County, North Dakota, USA, on 17-18, 1995. Eggs were delivered double
bagged, in double coolers, with ice packs between the inner and outer cooler,
by air to Yankton Field Research Station, Yankton, South Dakota, on May 19,
1995, for testing. The eggs were removed from plant detritus and placed in a
Heath incubator at 13°C. Well water quality for egg incubation was pH 7.59,
hardness as CaCO3 1,000 mg/L and alkalinity as CaCO3 286
mg/L. Larval culture well water quality was pH 7.51 - 8.27, hardness as CaCO3
770 - 1,050 mg/L and alkalinity as CaCO3 268 - 283 mg/L.
Dilution Water
Tests were conducted in American Society for Testing and Materials (ASTM 1989)
very hard reconstituted water (hardness and alkalinity as CaCO3 280
- 320 mg/L and 225 - 245 mg/L, respectively, and pH 8.0 - 8.4). Reconstituted
water was prepared by adding calcium sulfate, magnesium sulfate, potassium chloride,
and sodium bicarbonate to deionized water in polyethylene tanks equipped with
a recirculating pump to mix and aerate the water (Hamilton et al. 1989). Test
water was not prepared with reagent grade salts, as suggested by ASTM (1989),
because of the large volume of water needed. Water quality characteristics,
including chloride, calcium, sulfate, total hardness, and total alkalinity,
were measured for each tank of dilution water according to methods of American
Public Health Association (APHA 1989). Magnesium concentrations were calculated
using calcium and hardness values, conductivity (corrected to 25°C) was
measured with a Yellow Springs Instruments (YSI) Model 31 conductivity bridge
with a cell constant of K = 1.0, and pH was determined with an Orion
Model 901 Ionalyzer using a Ross combination electrode. (Use of trade names
throughout this article does not imply endorsement by the Federal Government.)
Test Chemical
The chemical used in the static renewal chronic test was atrazine (2-chloro-4-ethylamino-6-isopropylamine-s-triazine)
obtained from Ciba-Geigy Corporation, Greensboro, North Carolina. Test solutions
were prepared by pipetting appropriate aliquots of stock solution into 55-gallon
drums of reconstituted water. Stock solutions were prepared daily by dissolving
atrazine in reagent grade acetone then diluted with deionized water. Control
solutions were prepared by pipetting appropriate elicits of 35% acetone stock
solution into a 55-gallon drum of reconstituted water. Control and test solutions
were thoroughly mixed prior to use.
Toxicity Testing
A static renewal test was conducted following procedures outlined by the ASTM
(1989). The test exposed 102 individually housed salamander larvae to two concentrations
of atrazine, 75 or 250 µg/L, plus a control for 86 days (n = 34
for each treatment). An additional 102 larvae were treated in the same way,
but plasma from these larvae was used in a separate analysis which will be reported
elsewhere. We used all 204 larvae in growth and metamorphic timing analyses.
Larvae were assigned to one of two groups: those that would be sampled at stage
2 (early) and those that would be sampled at stage 4 (middle) of metamorphic
climax (Norman 1985). Larvae were tested individually in 7.6-L polyethylene
containers covered with screened lids, containing 4.0 L of test solution. Tests
were initiated when larval snout to vent length was 30 mm (
= 32.58 ± 0.46 mm). Larvae were allowed to gradually acclimate to test
water over a period of 2 d before atrazine was introduced; test water was renewed
daily during acclimation. After acclimation, test solution was renewed every
second day by removing one-half of the solution and adding an equal volume of
fresh solution. Larvae were fed three red worm pieces daily, supplemented with
brine shrimp naupii. Mortality was recorded daily and metamorphic stages were
determined using tail-fin measurements as described by Norman (1985). Upper
and lower tail-fin measurements were taken every third day until the animal
neared the next metamorphic stage, at which time measurements were taken daily.
Salamanders were killed and blood samples were collected in capillary tubes
and centrifuged; then plasma stored frozen at -25°C until analyzed. Before
sacrifice each salamander was weighed and measured; plasma corticosterone measures
obtained in this fashion cannot be considered basal, but because all larvae
were handled in the same way, we consider the concentrations to be comparable
among treatment groups.
Ambient test temperature ranged from 19.0° to 26.0°C. Water quality
parameters (see "Dilution water" above for list) in exposure vessels for both
studies were measured weekly during the study. Dissolved oxygen was measured
weekly in exposure vessels before and after renewal with a YSI Model 58 oxygen
meter. The photoperiod simulated conditions in North Dakota during mid-June
to mid-July.
Toxicant Analysis
Test water was analyzed to verify the concentration of atrazine in the test
vessels. Subsurface grab samples were collected in acid-cleaned 16 x 125 mm
glass culture tubes with Teflon-lined screw caps from 12 arbitrarily chosen
test vessels at day 1, 7, 15, 28, 42, 56, 70, and 84, and refrigerated until
analysis. Samples were analyzed in triplicate utilizing Ohmicron's Enyzme-Linked
Immunosorbent Assay tests for atrazine.
Radioimmunoassay for Corticosterone and Thyroxine
Plasma concentrations of corticosterone were assayed with techniques modified
from the method of Wingfield and Farner (1975) as reported in Fivizzani et al.
(1986) and Gratto-Trevor et al. (1991). Variations of these methods have been
used to quantify plasma steroid hormones in different vertebrates from fish
to mammals. Plasma samples (100-200 µl) were brought up to a final volume
of 400 µl with distilled water, and corticosterone was extracted with 5
ml of dichloromethane. The organic phase was removed and dried under a stream
of nitrogen. The extraction efficiency for this method in previous studies has
been between 80% and 95%.
The extract was suspended overnight in phosphate-buffered saline-gelatin
assay buffer. Duplicate 200-µl aliquots of the resuspended extract were
assayed with a highly specific antibody from Endocrine Sciences, Tarzana,
California (B3-163), and radioactive corticosterone (1,2,6,7 -3H)
from New England Nuclear, Boston. Chromatographic separation of the samples
was not performed as this antibody is highly specific with a cross-reactivity
of 3.3% with desoxycorticosterone and 1% or less with all other steroids.
To validate the use of this assay for Ambystome tigrinum, a pooled
sample of plasma from this species was serially diluted and was found to be
parallel to the standard curve. Additionally, a plasma volume from A. tigrinum
was stripped of endogenous steroids via multiple charcoal treatments, and
exogenous corticosterone was added in a known concentration to verify the
accuracy of the assay. The assay continued overnight at 4°C, and dextran-coated
charcoal followed by centrifugation was employed to separate the bound from
the free fractions. Aliquots of the bound fraction were counted on a Beckman
LS-6800 liquid scintillation counter. Plasma values were compared against
a series of corticosterone standards from Sigma Chemical Company. These methods
have been previously determined to allow for detection of corticosterone concentrations
as low as 1.5 pg/assay tube (between 30 and 60 pg/mL plasma depending on plasma
sample size). The intra-assay coefficient of variation was 8.7%.
Plasma thyroid hormone concentrations were determined according to the methods
discussed in Hoffnagle and Fivizzani (1990) utilizing solid-phase radioimmunoassay
kits (Micromedic Systems, Horsham, Pa.). The cross-reactivity of this assay
system was 1.3% with triiodothyronine and less than 0.1% with other similar
amino acids. All samples were analyzed in a single assay to eliminate the
problem of interassay variability. Standards were prepared by adding known
quantities of plasma thyroxine to tiger salamander plasma previously charcoal-stripped
of all endogenous thyroid hormones. Duplicate 5-µl undiluted samples
and standards were assayed, and the bound fraction counted with a Beckman
5500 gamma counter. Estimates of sample plasma thyroxine content were determined
via a comparison of each sample with a graph of the complete series of standards.
Statistical Analysis
We used ANOVA techniques to assess the effects of atrazine concentration (0,
75, or 250 µg/L), stage (2 or 4), and their interaction on plasma corticosterone
concentration, plasma thyroxin concentration, days-to-stage, weight, and snout-vent
length of larval tiger salamanders. Because the salamanders were allocated to
individual containers and then randomly assigned to one of the six stage-by-dose
combinations, we considered each individual salamander to be an independent
replicate and conducted the ANOVA as a simple two-way factorial (Kirk 1982).
We tested for differences in the change in the response variable from stage
2 to stage 4 among the three atrazine treatments by using specified contrasts
statements (Kirk 1982). We isolated differences among least squares means
(SAS 1989) for significant main effects and interactions using Fisher's protected
least significant difference value (Milliken and Johnson 1984). We used the
general linear models procedure of SAS (1989) to conduct the ANOVA analyses
and isolate differences among least squares means for significant effects
in the ANOVA, and the ESTIMATE statement for conducting the specified contrasts.
All statistical tests were conducted at the 0.05 significance level unless
stated otherwise.
Previous Section -- Introduction
Return to Contents
Next Section -- Results
U.S. Department of the Interior |
U.S. Geological Survey