Determination
of Nitrate
in Waters from
Fish Ponds
Leaflet 99/December 1980
CLAUDE E. BOYD and
WILLIAM D. HOLLERMAN*
Thbe P'H~NI!' \0 ISI ONIG A( 11) M O
hasbceA~icl\use fodetermiiningntae
nitrogen in wsater. In this proceduie. "rater is
esaporated and nitrate in the residue is
treated wsith phenoldisulfonic acid to form a
colorless nitiro dens ratis e. In alkaline solu-
tion, the nitro dens atis e is transformed to a
velloss -colo)red compouind. The intensits of'
the velloss color is pr oportional to the con-
centration 
of nitrate, 
permitting 
colorimetrN 
i rcdr 
stmecnuigbcueo
the es apotation step and because a number
of substances interlci e w.ith the anal\ sis.
Recentlx. another nitrate procedure has
gained popularitv. Nitrate is reduced quanti-
tatis el\, to nitrite \,,,hen a samrple is passed
over cadmium filings that base been treated
witth copper sulfate. 1 he nitrite is reacted
ssith sulfanilamide and N-( I -naphth\ I)-
eths lenediamine to f orm a highly colored
a/o dNxc so its concentration masr be mea-
sured colorimetricallsr. A correction must be
made for nit rite initial ly p resent in the
sample.
I he present studsr was initiated to compare
the precision and accuracv of' the twxo
methods of' determining nitrate in wxaters
from fish ponds.
* Pi Oessor and Reseat ch Associate, respec-
tisel\, Deparitment of F-isheries and Allied
ii aeul Itir es.
F-IG. 1. Cadmium reduction colfumn packed \Nith
cadmium filings.
METHODS
-The phenoldisulf'onic acid method ssas
conducted according to standard proce-
durel The cadmium reduction method fol-
lowsed the suggestions of Str ickland and
Parson S . Howsever, there wsere tsso Asaria-
tions from their suggestions. Instead of con-
structing the reduction columns from simple
Bo) in C. F. 1979. 
W\ater Qualit\ 
in Warm-
wsater Fish ponds. Auburn Umn (Ala.) Agr Exp.
ta, 359 p.
2 Si kin Ki \\i)_J D. L 
NDF R. Pszso'ss 1972. 
A
practical Handbook of Sea'sater '.nalssis. Bull.
67. Fish. Res. Bed. Canada. 311 p.
AGRICULTURAL EXPERIMENT STATION/AUIBURN UNIVERSITY
Gale A. Buchanan, Director Auburn University, 
Alabama
f~.A
materials, pre-fabricated glass columns, fig-
ure 1, were purchased from the Robert And-
erson Glass Company, Old Turnpike Road,
Fitzwilliams, New Hampshire. Cadmium fil-
ings were prepared according to Strickland
and Parsons, but before transferring the cad-
mium filings to the column, the filings were
washed several times with distilled water to
remove excess colloidal copper. Following
reduction of nitrate to nitrite in the column,
nitrite was determined by diazotization'.
Water samples were obtained from ponds
on the Fisheries Research Unit, Auburn Uni-
versity Agricultural Experiment Station,
Auburn University, Alabama. These samples
initially contained 0.2 mg/liter or less of ni-
trate-nitrogen, so small amounts of sodium
nitrate were added to some samples to in-
crease nitrate-nitrogen concentrations. Be-
fore analysis, all samples were filtered
through glass fiber filters (Gelman Type
A-E) to remove particulate matter.
Precision estimates were based on seven
replicate analyses of each of six samples
3
.
Percentage recovery was used to estimate
accuracy
3
. After determining nitrate initally
present, 28 samples were spiked with either
0.05 mg/liter nitrate-nitrogen (cadmium
reduction method) or 0.3 mg/liter nitrate-
nitrogen (phenoldisulfonic acid method) and
nitrate-nitrogen was again measured. The
percentage recovery was estimated by the fol-
lowing equation4
Percent recovery = x 100;
F = the final concentration of nitrate-
nitrogen; I = the initial concentration of
nitrate-nitrogen; S - the concentration of
nitrate-nitrogen added. The closer to 100
percent the recovery, the greater the accu-
racy. Finally, nitrate-nitrogen was deter-
mined in 25 samples by both methods and the
results were subjected to regression analysis.
3UNITED STATES ENVIRONMENTAL PROTECTION
AGENCY. 1972. Handbook For Analytical Quality
Control in Water and Wastewater Laboratories.
Analytical Quality Control Laboratory, Cincin-
nati, Ohio. 97 p.
4
BOYD, C. E. 1979. Determination of Total
Ammonia Nitrogen and Chemical Oxygen
Demand in Fish Culture Systems. Trans. Amer.
Fish. Soc. 108: pp. 314-319.
RESULTS AND DISCUSSION
The cadmium reduction method permits
determination of lower concentrations of
nitrate-nitrogen than are possible with the
phenoldisulfonic acid method, table 1.
Nitrate-nitrogen concentrations in fish
ponds are often below 0.1 mg/liter', so the
cadmium reduction procedure has a distinct
advantage over the phenoldisulfonic acid
method.
TABLE 1. COMPARISON OF ABSORBANCE
READINGS FOR NITRATE-NITROGEN
CONCENTRATIONS USING THE CADMIUM
REDUCTION METHOD AND THE
PHENOLDISULFONIC ACID TECHNIQUE
Absorbance in 1-cm
colorimeter cell*
Nitrate-nitrogen Cadmium Phenoldisulfonic
(mg/ liter) reduction acid
0.01 0.036 0.004
0.05 0.167 0.022
0.10 0.328 0.046
0.15 0.502 0.071
0.20 0.658 0.092
*Suitable accuracy for routine purposes can
be achieved if absorbance values fall between
0.025 and 0.700. For rigorous work, absorbance
values should be between 0.100 and 0.450.
Precision estimates, table 2, suggest that
both methods of nitrate analysis are rela-
tively precise. However, the precision of both
methods tends to decrease at low nitrate-
nitrogen concentrations. The percentage re-
covery of nitrate-nitrogen was significantly
greater (t = 5.14; P<0.001) for the cadmium
reduction method than for the phenoldisul-
fonic acid method, table 3, suggesting that
the cadmium reduction method is more accu-
rate. Percentage recovery for the phenoldi-
sulfonic acid method was even poorer
(58.3 percent) than indicated in table 3 for
eight samples initially containing less than
0.15 mg/liter nitrate-nitrogen. However,
even when these samples were removed, the
percentage recovery for the remaining 20
samples was only 91.6 percent. Percentage
recovery by the cadmium reduction method
was not affected by the initial nitrate-
nitrogen concentration.
Correlations between concentrations
determined in the same samples by the two
methods were not significant (r2 - 0.17; P>
0.05) for 13 samples containing low concen-
trations of nitrate-nitrogen. Since accuracy
[2]
TABLE 2. PRECISION ESTIMATES FOR NITRATE-NITROGEN DETERMINATIONS MADE BY THE CADMIUM
REDUCTION METHOD AND BY THE PHENOLDISULFONIC ACID TECHNIQUE. SAMPLES
WERE ANALYZED SEVEN TIMES
Sample'
A B C D E F
Cadmium reduction:
Mean (mg/liter) ............... 0.001 0.008 0.114 0.198 0.662 1.14
Standard deviation (mg/liter) .... 0.0004 0.0008 0.0036 0.0040 0.0042 0.010
Coefficient of variation (pct.) .... 25.0 10.0 3.2 2.0 0.6 0.9
Phenoldisulfonic acid:
Mean (mg/liter) ............... 0.03 0.15 0.21 0.61 0.81 2.93
Standard deviation (mg/liter).... 0.004 0.01 0.011 0.034 0.05 0.063
Coefficient of variation (pct.) .... 13.3 6.7 5.2 5.6 6.2 2.2
1
Sample A for cadmium reduction was not the same sample as Sample A for phenoldisulfonic
acid, etc.
TABLE 3. SPIKE-RECOVERY ESTIMATES
OF ACCURACY FOR NITRATE-NITROGEN
DETERMINATIONS MADE BY THE CADMIUM
REDUCTION METHOD AND THE
PHENOLDISULFONIC ACID TECHNIQUE
(TWENTY-EIGHT SPIKE-RECOVERY TESTS
WERE MADE WITH EACH PROCEDURE)
Cadmium 
Phenoldi-
reduction 
sulfonic
acid
Pet. Pet.
Mean ............... 99.6 82.1
Standard deviation ... 2.96 17.8
Coefficient of variation . 3.0 21.7
was superior for the cadmium reduction
method, table 3, data obtained by the phe-
noldisulfonic acid method were assumed to
be erroneous. There was a significant correla-
tion (P< 0.01) between nitrate-nitrogen con-
centrations determined on the same samples
for higher nitrate-nitrogen levels, figure 2.
However, the cadmium reduction method
gave consistently greater values than the
other method.
CONCLUSIONS
Because of its greater sensitivity and accu-
racy, the cadmium reduction method is
superior to the phenoldisulfonic acid tech-
nique for measuring nitrate-nitrogen in
waters from fish ponds. The cadmium reduc-
tion columns are initially difficult to prepare,
but once columns are prepared, the cadmium
reduction method is no more complicated to
perform than is the phenoldisulfonic acid
procedure.
0.5-
-0 0
O I I I I I I I I I a a
0 0.5 1.0
CQdmium reduction N0
3
-N (mg/liter)
FIG. 2. Relationship between nitrate-nitrogen
concentrations measured on thesame samples by
the cadmium reduction method and the phenoldi-
sulfonic acid method.
[3 ]
Phenoldisulfonic
ocid N0
3
-N (mg/liter)
1.0 = :-0.044+ (
r = 0.93
0.771X
Information contained herein is available to all without
regard to race, color, sex, or national origin.
[4 ]