~I7~i~
C
Leaflet 105 R. Wu and C.E. Boyd' December 1989
DISSOLVED OXYGEN (DO) is a
critical variable in the management of
aquaculture ponds, and fish farmers must
have a reliable means of measuring its con-
centration. Inexpensive water analysis kits
can be used to obtain accurate measurements
of DO, but the procedure is time consum-
ing and difficult to conduct at night when
DO concentrations are most critical.
Therefore, fish farmers with several ponds
need a reliable instrument that permits rapid
measurements of DO and which can be
operated with ease in the dark. Portable DO
meters appear to meet this need, provided
they are accurate. Several brands of high
quality DO meters are available at a cost of
$750 to $1,250 each. Recently, a few com-
Visiting Scientist and Professor, respectively,
of Fisheries and Allied Aquacutures.
' Boyd, C. F. 1976. An Evaluation of a Water
Analysis Kit. Leaflet 92, Ala. Agr. Exp. Sta.,
Auburn Univ.
panies have been offering less expensive DO
meters ($250 to $350) for use in aquaculture.
Fish farmers sometimes need data on pH
of water in hatcheries or ponds. Again, water
analysis kits can be used, but color tests
employed in kits usually provide pH readings
0.5 to 1.0 pH unit too high.' Conventional
laboratory pH meters cost roughly the same
as high quality DO meters. However, pocket
pH meters are now available at low cost (less
than $50) for use by fish farmers.
DESCRIPTION OF TEST
To determine the accuracy of DO meters,
a study was conducted in 1987 and 1988 at
the Alabama Agricultural Experiment Sta-
tion. Comparisons were made of DO con-
centrations measured with two inexpensive
DO meters (Digital Oxygen Analyzer,
Engineered Systems and Designs,
Wilmington, Delaware, and Sentry III
Alabama Agricultural Experiment Station Auburn University
Lowell T. Frobish, Director Auburn University, Alabama
? . .. ..... . &.. ~.~l. ~ ii~ L.. ~. - ..sMrj .t
Oxygen/Temperature Monitor, Barebo,
Inc., Emmaus, Pennsylvania) and a pocket
pH meter (Hach Company, Loveland, Col-
orado). Values obtained with these in-
struments were compared with values ob-
tained with a conventional, portable DO
meter and a standard, line-operated,
laboratory pH meter.
During 1988, it was necessary to make DO
measurements in a number of research ponds
at the Fisheries Research Unit. Therefore, on
several occasions between March and Octo-
ber 1988, probes of the three DO meters were
tied together with a rubber band, and DO
concentrations were measured at the same
time and place with all three instruments. To
compare pH measurements by the two
meters, water samples were brought into the
laboratory for pH analysis.
RESULTS
Concentrations of DO ranged from less
than 1 p.p.m. (part per million) to 14 p.p.m.
and water temperatures were from 50 to
86 'F. Over this broad temperature and DO
range, DO estimates from the two in-
expensive meters were highly correlated with
concentrations determined with the conven-
tional DO meter, figure 1. The conventional
DO meter was found to repeatedly read the
correct DO concentrations in prepared
samples, so DO concentrations obtained by
the conventional meter were considered true
DO values. The regression equations, figure
1, were used to calculate the average DO con-
centration measured by the two meters at dif-
ferent true DO concentrations. Relative
errors and differences between measured
and true values are provided below:
DifRelative
True DO, Dference error, pct.
p.p.m. Digital Sentry itaSentryI1 rDiaitl 
enry
1 .............- 0.63
2 ............- .36
3 ............ - .09
4............ + .18
5 ........... .+ .45
6 ............ + .72
7 ............ + .99
8............ + 1.12
10 ........... + 1.80
12........... + 2.34
- 0.35
- .14
+ .07
+ .28
+ .49
+ .70
+ .91
+1.26
+1.15
+ 1.96
- 65.0
- 18.0
- 3.0
+ 4.5
+ 9.0
+ 12.0
+14.1
+15.8
+11.8
+ 19.5
III
- 37.0
S14.0
+ 2.3
+ 7.0
+ 9.8
+ 13.0
+14.0
+ 15.4
+ 16.3
Percent relative error =
100 x true value - measured value.
true value
Measured D.O., p.p.m. (Y)
16 - Y =-0.90 
+ 1.27X
14 - .
12
. o
8 -
8
6 - *. Digital
4 .'s
2
10 ..
1 2 - s
0 2 4 6 8 10 12 14
True DO., p.p.m. (X)
ygen (DO) concentrations measured with a
Digital or Sentry I DO meters (measuredII
0
True D.O., values, Y). 
(X)
FIG. 1. Regative sign betwfore a dissolvfference or ax-
relative error indicates that a DO meter gave
a lower reading than the true value, while a
positive sign indicates that a DO meter
overestimated the true value. At low DO con-
centration, both the Digital and Sentry III
meters underestimated true DO. When DO
was 4 p.p.m. and above, the instruments
overestimated true DO concentrations.
Nevertheless, both meters provided a good
approximation of DO concentration. When
DO is low, management errors would not
result:from underestimations in DO of the
magnitude found here. When DO concentra-
[2]
Measured pH (Y)
12
Y = -0.47 + 1.05X
10 -.
8 .
6
4
2
0
0 2 4 6 8 10 12
True pH (X)
FIG. 2. Regression between pH values
measured with a laboratory pH meter (true
values, X) and an inexpensive, Hach pocket
pH meter (measured value, Y).
tions are above 4 p.p.m., there is adequate
DO for warmwater fish and overestimations
of magnitudes observed would not be
serious.
Durability of the two DO meters is largely
unknown. The Digital meter was used for 2
months in 1987 and during the entire 1988
growing season with no malfunctions ob-
served. The Sentry III meter was used dur-
ing the 1988 growing season, and no
malfunctions occurred. Conventional DO
meters are more expensive and of better
quality than the inexpensive meters tested.
Several conventional DO meters have been
used in water quality research by the Depart-
ment of Fisheries and Allied Aquacultures.
These meters have functioned properly for
3 to 6 years before repairs were necessary.
The pocket pH meter also provided pH
values highly correlated with measurements
made with the laboratory pH meter, figure
2. Using measurements made with the
laboratory pH meter as true values, relative
errors and differences between values
measured with the pocket pH meter and true
values were:
True pH
3 .... .......
4 ...........
5 .... .......
6 .. .........
7 ...........
8 ... . .......
9 . .. . .. .....
10 ... .......
Difference
-0.32
- .27
- .22
- .17
- .12
- .07
- .02
+ .03
Relative
error, pct
- 10.7
- 6.8
- 4.4
- 2.8
- 1.7
- .9
- .2
+ .03
Agreement between measured and true
values improved as pH increased. The pocket
pH meter provided sufficiently accurate
measurements for use in practical
aquaculture. It is not known how long the
meter will last, but it did not malfunction
during the 1988 growing season.
Information contained herein is available to all
without regard to race, color, sex, or national origin.
[3]