United States
Environmental Protection
Agency
Robert S. Kerr Environmental Research „
Laboratory X/F, vs
Ada OK 74820
Research and Development
EPA-600/S2-83-019 May 1983
SERA Project Summary
An Evaluation of Filter Feeding
Fishes for Removing Excessive
Nutrients and Algae from
Wastewater
Scott Henderson
This study was instituted to
determine the feasibility of utilizing
certain species of f inf ish for the removal
and recycling of excessive nutrients and
algae from wastewater. The silver carp
(Hypopthalmichthyes molitrix) and the
bighead carp (Aristichthyes nobilis)
were chosen as the central species due
to their specifically adapted filter
feeding mechanism. An existing
wastewater treatment plant with six
lagoons served as the project site. Since
the results from previous controlled
field trials were available, this project
utilized the entire facility as a pilot scale
system. No attempt was made to alter
or influence the waste load normally
received by the lagoons.
It can be said unequivocally that the
presence of the fish had a beneficial
effect on the aquatic system. Because
of the many variables involved in such a
dynamic, stressed ecosystem it is
difficult, if not impossible, to quantify a
direct relationship between the
standing crop of fish and any one water
quality parameter. In all. 14 water
quality parameters along with selected
heavy metals, pesticides, pathogenic
bacteria, and viruses were monitored
during the project.
Analysis of the data shows that the
presence of the fish improves the
treatment capability of the
conventional lagoon system. There are
tradeoffs to be made among some
parameters and some liabilities
resulting from the presence of the fish.
All are within acceptable limits and.
when considered, still tip the scales in
favor of the benefits gained. In the final
analysis, the real determining factor in
deciding to use a finfish-aquaculture-
treatment system is the capability of
using the more than 7,200 kg/ha
annual production of fish as a revenue
producer to sufficiently offset or pay for
water treatment costs.
This Project Summary was developed
by EPA's Robert S. Kerr Environmental
Research Laboratory. Ada. OK. to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
The passage of the Federal Water
Pollution Control Act of 1972 (PL 92-500)
generated considerable interest and
concern for the development of
wastewater treatment methods that
would meet the more stringent standards
at a reasonable cost. The emphasis on
re-use of wastewater and the recycling of
nutrients into useful products brought
about a new look at old biological treat-
ment methods. The biological production
capability of nutrient-laden wastewaters
is obvious. However, directing the energy
and raw materials into useful products
has proven difficult.
Often the emphasis has been on
developing new products and uses for the
mostly invertebrate species that grow
naturally in wastewater. With an already
growing demand and a decreasing world
supply of fish and fisheries products.
-------
many investigators have attempted to
rear fish in wastewaters. This has been
largely unsuccessful in the United States
due to the lack of a native species with a
high production capability utilizing
primary production from ponds or lagoons
as a food source. The importation of the
silver and bighead carp into Arkansas in
1973 by a private fish farmer provided the
opportunity for experimentation with fish
species uniquely adapted for the job.
Initial interest in the silver and bighead
carp resulted from an extensive amount
of literature reporting the many
characteristics they possess that make
them a seemingly ideal fish for culture. A
fish that could be added to the species in
Arkansas' large fish farming industry to
increase production was an attractive
possibility. It became apparent that these
filter feeders had quite an impact on
water quality and this became an
increasingly important subject of
subsequent studies. All preliminary
work corroborated reports in the litera-
ture concerning production and growth
rate potential of these fish. By the time
this project was designed and
implemented, the major emphasis was
on the use of these fish to improve the
quality of wastewater. This was
somewhat unique in that previous work
had been concerned with the optimum
nutrient loads to add to ponds to
maximize their production. The ability of
the fish to withstand heavy wastewater
loads and their concomitant impact on
water quality is relatively unexplored
territory.
Fertilization of fish ponds has long been
recognized by the fish culturist as a
method of increasing production. The
production of finfish as a method of
reducing fertility is a relatively new
approach that has been stimulated by the
increasing need for effective, low cost
treatment of wastewater by small
municipalities. The recent realization of
the need to conserve energy and to
recycle what has previously been re-
garded as a troublesome waste prpduct
has provided the impetus for exploring
alternative methods. The Arkansas Game
and Fish Commission's interest in this
project evolved from the fact that they
were an unknown, exotic species being
imported into the state and the need to
evaluate the possible dangers as well as
the beneficial uses of the fish.
The fact that X pounds of fish are
produced without supplemental feeding
obviously shows that in one fashion or
another, energy and nutrients are
transformed into the very stable form of
fish flesh. The fish culturist may draw on
a rather large body of literature in deter-
mining the proper type and amount of
fertilizer to add to a culture pond. If, on the
other hand, the objective is to utilize
available nutrients, little is known about
the effectiveness of finfish in general or
any species in particular. Common sense
dictates that those fishes that have
adapted to feeding at the lower trophic
levels would be most efficient at
converting nutrients. The filter feeding
silver and bighead carps meet these
criterion and are the key species in this
study.
Project Site
An existing wastewater treatment
plant of the Benton Services Center in
Benton, Arkansas was chosen as the
location for the study. Other than the
collective individual wastes of approxi-
mately 1,000 residents and employees of
the center, the largest contributors to the
center's treatment plant are a laundry
and full-time food services section.
The physical facilities of the treatment
plant are the same as many small
municipalities in Arkansas and consist of
(1) bar screen and grinder, (2) clarifier, (3)
aerobic digester, and (4) six lagoons.
Minor alterations were made to the
plant before initiating the study, but for
the most part it functioned on a day-to-
day basis the same as other lagoon type
treatment facilities. At no time was the
normal wastewater load coming to the
plant altered during the two year project
period.
Monitoring Program
Water Quality
Grab samples were taken regularly at
the effluent of each of the six ponds. All
sampling was done according to the
APHA Standard Methods with analysis
being conducted by the Arkansas
Department of Pollution Control, and
Ecology. Parameters monitored were:
Air Temperature
H20 Temperature
Carbon dioxide
Dissolved oxygen
BODj
Turbidity
Ammonia
Nitrite - N
Nitrate - N
pH
Total Suspended Solids
Total Phosphorus
Fecal Coliform
Plankton Enumeration
Toxic Substances
Due to the need to utilize the fish to
provide an economic return as well as
maintain an expanding population for
optimum efficiency, samples were taken
to look for the presumed most likely toxic
substances. These samples of both fish
and water were analyzed by an
independent testing laboratory for:
Aldrin
Dieldrin
Endrin
Mi rex
DDT
Toxaphene
Kepone
PCB
Lead
Copper
Cadmium
Mercury
Arsenic
Biological Contaminants
Since human consumption is
considered the ultimate use of the fish
produced, the Baylor College of Medicine
was contracted to provide analysis of
bacterial and viral contamination.
Regular samples were taken of the water
and bottom mud as well as various
portions of the fish. All samples were
assayed for common viral and bacterial
pathogens associated with municipal
wastewaters.
Fish Production
To monitor the growth rate of fish
within the system, monthly samples were
taken throughout the growing season
and individual fish weighed, measured
and returned to the ponds. It was difficult
to obtain adequate samples of species
other than the silver carp due to relatively
low stocking densities and the inefficien-
cy of sampling techniques in the 1.5-1.8
hectare ponds.
Ponds 1 and 2 were considered to be
plankton culture ponds necessary to
accept the initial shock of the BOD loads
and stabilize dissolved oxygen levels. As
would be expected, the serial
arrangement of the ponds provided
successively better water quality in each
successive pond. Pond 3 was extremely
fertile with a heavy plankton bloom and
typically minimum dissolved oxygen
levels. Pond 4 exhibited wide fluctuations
in DO levels but other water quality
parameters began to stabilize over time.
Ponds 5 and 6 remained in near optimum
conditions for pond fish culture
throughout the project period.
-------
Economic Design Considerations
Since the project was conducted in a
full scale functioning sewage treatment
system, the experience provided the
ability to make certain realistic
projections regarding economic and
design considerations. This type of finfish
wastewater treatment system has shown
the capability of upgrading the effluent of
conventionally designed and operated
lagoon treatment plants. However, the
level of treatment is somewhat limited
compared with other types of advanced
treatment systems. Only when the fish
produced from recycling the wastes can
be utilized, can the true advantages of this
method of treatment be realized. If a true
profit or even supplemental income to
offset treatment costs can be generated,
the production of fish becomes a more
attractive treatment method or a viable
addition to more advanced treatment
practices.
Results and Discussion
Water Quality
The wastewater entering the plant had
an average BOD5 of 251.4 mg/l and a
suspended solids concentration
averaging 97 mg/l. During the two year
project period, the system has reduced
the BOD5 by 96.01% and the TSS by
78.22%. Also, the effluent has been
within the criteria established for second-
ary wastewater treatment and many
parameters were at levels 'associated
with advanced secondary treatment.
Toxic Substances
With the exception of the metals,
copper and mercury, all samples have
contained less than the standard
detection limits or have been negative. In
no instance has any sample contained
the listed contaminants at levels above
action guidelines established by the FDA
or the Arkansas Department of Health.
Biological Contaminants
From the influent to Pond 6, there was
an average 2.6-fold decrease per pond for
fecal coliforms (FC) and an average 2.4-
fold decrease per pond for fecal
streptococci (FS). Bacterial concentra-
tions in the sediments followed a
different pattern than in the overlying
waters. There was an average 4.7-fold
decrease per pond for FC in the last four
ponds. The decrease in FS in the sedi-
ments from pond to pond was substan-
tially less than the decrease in FS in the
water.
The concentrations of FC and FS in the
fish guts were on the average greater
than in the surrounding water and
sediments. Mean concentrations of FC
and FS in the fish skin were lower than in
the gut. Levels of bacteria detected in the
fish flesh are shown in Table 1.
Two methods were used for sampling
the fish muscle. The samples in August
and September were taken by a normal
fillet procedure using a decontaminated
fillet knife. Samples taken during these
two months yielded sporadically high
levels of FC and FS in the muscle tissue,
probably due to contamination by
bacteria from the fish skin. Beginning in
October, all muscle samples were taken
aseptically to avoid contamination from
the skin. Three of nine samples of muscle
tissue obtained from October through
December were positive for either FC or
FS at low levels.
Salmonella spp. was detected in 2 of 4
influent samples at levels of 0.4 and 2.3
MPN (most probable number)/100 ml
using dulcitol selenite enrichment.
Salmonella spp. was also isolated from a
single water sample from Pond 2 in
December. No salmonella was isolated
from any of the other pond water,
sediment, or fish samples.
Six of the 90 samples tested for enteric
viruses yielded at least 1 PFU (plaque-
forming unit) on BGM (buffalo green
monkey kidney cells) monolayers (Table
2). Three of five influent samples were
positive at low levels with concentrations
ranging from 7.5 to 20 PFU/liter. Two of
15 sediment samples and one water
sample from Pond 2 also yielded 1 -2 PFU
per sample of 500 g or 20 liters. All other
pond water samples were negative for
virus. No viruses were detected in any of
the 45 fish samples processed.
The sewage entering the Benton fish
ponds was atypical from a virological
standpoint. The levels of virus in the
sewage were much lower than would be
expected for untreated sewage from a
larger and more diverse community. For
example, concentrations in raw sewage
from treatment plants in St. Petersburg,
Florida, averaged 90 PFU/liter and, at a
larger treatment plant in Tampa, Florida,
concentrations of over 2,000 PFU/liter
were found. The sewage entering the
Benton ponds had an average
concentration of <9 PFU/liter for the 5
samples tested.
Because of the low levels of virus found
in the influent, the results cannot be
extrapolated to make conclusions or
predictions about the survival and
transport of viruses in other fish pond
systems that may have a much higher
input of viruses. The lack of virus isolates
from the fish and pond water in this study
does not preclude the possibility of
viruses surviving in the fish ponds and
being accumulated by the fish if the initial
levels of virus were higher. In fact, since
relatively high levels of FC and FS were
found in the ponds and fish, it is likely that
viruses would also be present if the input
rate were higher, since viruses generally
survive inactivation processes better
than do indicator bacteria.
Fish Production
Ponds 1 and 2 were considered to be
plankton culture ponds necessary to
accept the initial shock of the BOD loads
and stabilize dissolved oxygen levels. Fish
were stocked in Ponds 3, 4, 5, and 6.
Other than the initial regrading of the
pond bottoms to facilitate harvesting, no
supplemental aeration or fresh water
was provided to any of the fish ponds. All
were left in series accepting the full flow
volume and waste load as it passed
through the plant. As long as the entire
system functioned normally, all four of
the fish ponds maintained adequate
water quality for survival and growth.
As would be expected, the serial
arrangement of the ponds provided
increasingly better water quality in each
successive pond. Pond 3 was extremely
Table 1.
Concentrations of FC and FS in Fish Flesh
MPN/100 g fish flesh
Month
Aug.*
Sept.*
Oct.
Nov.
Dec.
Pond 4
FC
<30
230
<11
11
<11
FS
140
80
25
<11
<15
FC
<30
430
<11
<11
<11
Pond 5
FS
140
22,000
<11
<11
<15
Pond 6
FC
40
<30
<6.6
<11
<11
FS
860
<60
15
<11
<15
'August and September samples were taken by a normal fillet procedure with possible
contamination from the skin. All other flesh samples were taken aseptically.
-------
Table 2. List of Samples which Yielded
Plaque-Forming Units (PFUJ on
Cell Monolayers
Total PFU Estimated
Sample Month counted* concentration**
Table 3. Growth of Silver and Bighead Carps in Pond 5 During Project Period
Influent
Influent
Influent
Pond 2
sediment
Pond 4
sediment
Pond 2
water
Aug.
Nov.
Dec.
Sept.
Nov.
Dec.
5
3
2
2
1
1
20 PFU /liter
15 PFU /liter
7.5 PFU/liter
2 PFU/ 500 g
1 PFU/ 500 g
0.05 PFU/liter
* Unidentified.
** Dilution factors varied for influent samples.
fertile with a heavy plankton bloom and
typically minimum dissolved oxygen
levels. Pond 4 exhibited wide fluctuations
in DO levels and other water quality
parameters began to stabilize. Ponds 5
and 6 remained in near optimum
conditions for pond fish culture through-
out the project period.
In May of 1980 after the system had
been operational for 11/2 years, a delivery
line collapsed necessitating the flow of
the total raw waste load directly into Pond
2 until repairs could be made. In the six
weeks required for these repairs, the
already marginal water quality in Pond 3
deteriorated until a total oxygen depletion
and fish kill occurred on July 1, 1980.
Recovery of the fish from Pond 3 after the
kill showed that the original stocking
biomass of 374.8 kg/ha had increased to
7,165.1 kg/ha in the 18 months the fish
had been in the pond.
The plant breakdown and resulting
short-circuiting of the water also had a
visual impact on Pond 4. The period of
decreased retention timegreatly added to
the fertility in Pond 4. In essence. Pond 4
became Pond 3 during that period of time.
The diminished water quality coupled
with extremely hot, dry late summer
weather and a period of cloudy days
resulted in a major fish kill occurring in
this pond on September 4, 1980. A total
of 7,691.9 kg/ha of silver carp were
removed from Pond 4 as a result of this
kill. This is a considerable production in
the 21 months since initial stocking with
40.6 kg/ha.
The fish in Ponds 5 and 6 survived the
full 24 month experimental period.
Stocking and harvesting data for these
ponds are listed in Tables 3 and 4. Water
quality remained good and no problems
Date
Jan.. 1979
March, 1979
June, 1979
Sept.. 1979
Dec.. 1979
March, 1980
June. 1980
Sept.. 1980
Dec.. 1980
Time from
stocking
0
3 mos.
6 mos.
9 mos.
12 mos.
15 mos.
18 mos.
21 mos.
24 mos.
Silver carp.
standing crop
(kg/ha)
293.2
900.0
1,871.9
4.098.9
4,650.0
5,350.5
6.075.0
7.260.0
7.634.4
Bighead carp,
standing crop
(kg/ha)
39.1
425.6
560.0
1.510.4
Table 4. Growth of Silver and Bighead Carps in Pond 6 During Project Period
Date
Jan.. 1979
March, 1979
June, 1979
Sept.. 1979
Dec.. 1979
March, 1980
June, 1980
Sept.. 1980
Dec.. 1980
Time from
stocking
0
3 mos.
6 mos.
9 mos.
12 mos.
15 mos.
18 mos.
21 mos.
24 mos.
Silver carp,
standing crop
(kg/ha)
210.6
1,029.6
1,745.0
2.480.5
2.475.0
3.441.3
3.650.5
4,255.0
4.454.7*
Bighead carp,
standing crop
(kg/ha)
12.24
15.4
248.2
589.0
*Channel catfish, grass carp, and smallmouth buffalo were also initiallytstockedjnPond 6. Due to
low stocking rates, difficulty of sampling, etc., no interim growth estimates were made.
Also, the buffalo spawned during the spring of 1979 further complicating matters. At harvest, the
final standing crop for each species was found to be: channel catfish = 832 kg/ha.
buffalo = 562 kg/ha. grass carp = 262 kg/ha.
with the survival and growth of the fish
were noted.
Economic Considerations
This type finfish wastewater treatment
system has shown the capability of
upgrading the effluent of conventionally
designed and operated lagoon treatment
plants. However, the level of treatment is
somewhat limited compared with other
types of advanced treatment systems.
Only when the fish produced from
recycling the waste can be utilized, can
the true advantages of this method of
treatment be realized. If a true profit or
even supplemental income to offset
treatment costs can be generated, the
production of fish becomes a more
attractive treatment method or a viable
addition to more advanced treatment
practices.
Silver and bighead carp from a
preliminary hatchery study were
rendered into fish meal which assayed at
a crude protein content of a minimum 55-
57%. This is compared to 62% crude
protein for Menhaden meal considered
the best product now available. Oil and fat
content were not considered. There was
an estimated 18% return of meal from
fresh fish by weight. Current market
prices for pure fish meal, FOB Little Rock,
vary from $400-500 per ton in bulk
quantities depending on season and
harvest source. Based on a price of 7-9
cents per kg (3-4 cents per lb.)for live fish
and an annual production rate of approx-
imately 5,000 kg/ha as demonstrated in
this study, a gross return of $350-450 per
ha/yr could be realized by processing the
fish in this way.
If, on the other hand, human health
considerations could be mollified and the
product sold for direct human
consumption, the economic picture could
be quite different. Hatchery reared silver
and bighead carp have been tested organ-
oleptically for two different methods of
preparation. As a fresh fish fillet product,
the silver carp has a white, lightly oily
-------
meat that is excellent in a variety of
preparations with the subjective taste
test yielding comments ranging from
excellent flavor to barely acceptable.
However, the problem is boniness. The
silver carp has many floating bones that
do not increase in size proportionately as
the fish grows. This is a major problem for
American tastes even with larger sized
fish. Canning the fish, however, makes
the small bones unnoticeable and the
heat involved could overcome some of the
health effects problems. If the fish were
marketed in either manner (fresh or
canned), a conservative price of 55-65
cents per kg would be reasonable. The
gross amount return based on these
assumptions and the demonstrated
production potential would be $2,750-
3,250 per ha per year. Whatever the
market, any income realized would
certainly be welcomed to offset treatment
costs.
Design Considerations
In general, the factors involved in the
selection, design, and construction of a
finfish wastewater treatment system are
the same as those historically used for
conventional aerobic lagoon treatment
plants. Prime consideration should be
given to climate, availability of land area,
and the treatment level desired or
necessary. The results of this study have
shown that the additions of controlled
stocking of certain species and numbers
of fish can increase the efficiency of
lagoon treatment. Therefore, in instances
where conventional lagoon design criteria
indicate the system would be marginal,
either'due to space or treatment level, the
incorporation of finfish into the design
could make this the method of choice.
Since the fish must survive to do the job,
the most obvious criterion is that the
wastewater contain no contaminants
lethal to the organism. This could limit
use to specific circumstances or, more
likely, require in-house removal of these
substances prior to treatment. Because of
the flexibility needed to insure proper
operation, a finfish treatment system
requires a multiple lagoon design with
generally a serial flow pattern. The initial
impact of the BOD load from raw
wastewater must be lessened by some
method prior to entering the pond
containing fish. Short-term peaks in
loading rate are no major problem, but
generally the concentration of BOD5
entering the first pond containing fish
should be no more than 50 ppm annual
average.
There must be the capability of draining
each pond individually for maintenance
and harvest of the fish while allowing
continued operation of the plant. Typical
pond construction is applicable with the
probable need for a more carefully graded
bottom with a catch basin to facilitate
harvest of the fish. Little effect on water
quality is seen until the standing crop
of fish reaches 1,000 kg/ha. Also, larger
numbers of smaller, younger fish are
more efficient than fewer larger fish even
though biomass may be the same. A
method of harvest and replacement of the
fish should be established to maintain a
total standing crop between 1,000-5,000
kg/ha at all times and to have a high
percentage of small growing fish. Harvest
and restocking should be done annually
to provide maximum fish production or
should be done at least every three years
to prevent decreased water treatment
capability.
Conclusions and
Recommendations
The addition of silver and bigheadcarp
to a lagoon wastewater treatment system
increases the efficiency of that system.
Depending on' climatic and other
operational conditions, the inclusion of
these natural filters can increase
treatment levels by as much as 25-30%.
From a practical standpoint, this could
decrease the amount of land area needed
or improve the quality of water leaving
the facility, or both. When used as the
sole method of treatment, an aquaculture
system using silver carp is limited in
capability. Properly designed and
operated, the system could provide
advanced secondary treatment and
consistently meet discharge require-
ments of 10 ppm BODS and 20 ppm total
•suspended solids. Though nutrient
removal is improved and both total
phosphate and nitrogen levels were
decreased by more than 90% in this
system, total removal would require such
a lengthy retention time as to be
impractical. However, where finfish
treatment level requirements do not
exceed the capability of the system,
finfish aquaculture in wastewater
lagoons is a viable and reasonable
method of upgrading treatment and
recycling wastes into a stable and useful
form.
Aquaculture treatment systems are
competitive with other conventional
methods from a cost effectiveness
standpoint at the present time. Recycling
wastes into useful products is certainly
the ultimate goal of waste disposal. This
method achieves that goal in theory since
fishery products are in high demand. At
the present time, however, product
utilization possibilities range from being
limited to virtually impossible. The
development of quality control standards
to allow the use offish products grown in
wastewater is the most pressing need. If
that could be accomplished, there is little
doubt that a treatment system that could
potentially produce a profit would be
available.
Although the research described in this
article has been funded wholly or in part
by the U.S. Environmental Protection
Agency through cooperative agreement
number R805453 to the Arkansas Game
and Fish Commission, it has not been
subjected to the Agency's required peer
and policy review and therefore does not
necessarily reflect the views of the
Agency and no official endorsement
should be inferred.
. S. GOVERNMENT PRINTING OFFICE: 1983/659 -095/1939
------- |