Biological phosphate removal

Aquacultural Engineering 22 (2000) 121 – 136
www.elsevier.nl/locate/aqua-online

Biological phosphate removal in a prototype
recirculating aquaculture treatment system
Yoram Barak, Jaap van Rijn *
The Hebrew Uni6ersity of Jerusalem, Department of Animal Sciences, Faculty of Agricultural,
Food and En6ironmental Quality Sciences, PO Box 12, Reho6ot 76100, Israel

Abstract
Efforts to reduce phosphorus concentrations in aquaculture systems have mainly dealt
with improving the bioavailability of phosphorus in fish feed. Once released into the culture
water, phosphorus is generally left untreated and discharged with the effluent water. In the
present study, results are presented on a prototype recirculating treatment system originally
designed for removal of organic matter and inorganic nitrogen. Phosphorus determinations
in the various compartments of the treatment system (a digestion basin, a denitrifying
fluidized bed reactor and a nitrifying trickling filter) revealed that, after 210 days of
operation, more than 90% of the added phosphorus was retained within the organic matter
of the trickling filter. By means of batch experiments with bacterial consortia from the
reactors and with denitrifying isolates, it was found that denitrifiers were capable of
phosphate uptake in excess of their metabolic requirements. The phosphorus content of
organic material in the fluidized bed reactor was as high as 11.8% (on a dry-mass basis) while
it was much lower in the trickling filter (around 1.9%). Anoxic incubation of the trickling
filter material in the presence of an external carbon donor resulted in considerable denitrification activity and phosphate uptake. This finding served as an additional indication for the
fact that phosphate removal from the water in the system was mainly mediated by
denitrifying organisms. Based on these findings, the feasibility of using denitrification to
control phosphate levels in the culture and effluent water of recirculating aquaculture
systems is discussed. © 2000 Elsevier Science B.V. All rights reserved.
Keywords: Recirculating systems; Phosphate removal; Denitrification; Effluent treatment

* Corresponding author. Tel.: + 972-8-9481302; fax: + 972-8-9465763.

E-mail address: (J. van Rijn)
0144-8609/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 1 4 4 - 8 6 0 9 ( 0 0 ) 0 0 0 3 6 - 4


122

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

1. Introduction
Water quality deterioration due to excessive nutrient loading is of great concern
in intensive, recirculating fish culture systems. Often this concern not only relates to
the water quality requirements of the cultured animals but also to the quantity and
quality of waste discharge from these systems. The latter concerns arise from more
stringent environmental restrictions and concomitant higher levies on waste
discharge.
The principal sources of aquaculture wastes are uneaten feed and excreta. The
bulk of this waste is in the particulate form and in recirculating systems this is often
removed in a concentrated form by gravitational or mechanical methods (Chen et
al., 1994). Dissolved organic and inorganic nutrients, making up a smaller fraction
of the total waste in these systems, are removed with the effluent water. Among the
dissolved inorganic nutrients, nitrate, the end product of nitrification, is usually
present at high concentrations in the effluent of recirculating systems. Also phosphorus effluent concentrations are high due to the fact that much of the phosphorus
added with the feed is unutilized by the fish (Rodehutscord and Pfeffer, 1995) and,
in addition, due to the lack of appropriate methods for phosphorus removal in
these systems.
Enhanced biological phosphorus removal (EBPR) from domestic wastewater in
activated sludge plants is accomplished by alternate stages in which the sludge is
subjected to anaerobic and aerobic conditions. Under these conditions, phosphorus
is released from the bacterial biomass in the anaerobic stage and is assimilated by

these bacteria in excess as polyphosphate (poly-P) during the aerobic stage.
Phosphorus is subsequently removed from the process stream by harvesting a
fraction of the phosphorus-rich bacterial biomass (Toerien et al., 1990). Recently,
evidence was provided that some of these organisms are also capable of poly-P
accumulation under denitrifying conditions, i.e. with nitrate instead of oxygen
serving as the terminal electron acceptor (Barker and Dold, 1996; Mino et al.,
1998). As recently reviewed by Mino et al. (1998), studies on poly-P accumulating
organisms have revealed the involvement of specific set of metabolic properties
under anaerobic, aerobic and anoxic conditions. Under anaerobic conditions,
acetate or other low molecular organic compounds are converted to polyhydroxyalkanoates (PHA), poly-P and glycogen are degraded and phosphate is released.
Under aerobic and anoxic conditions, PHA is converted to glycogen, phosphate is
taken up and poly-P is intracellularly synthesized. Under the latter conditions,
growth and phosphate uptake is regulated by the energy released from the
breakdown of PHA.
In previous studies (van Rijn et al., 1995; Aboutboul et al., 1996) we showed that
discharge of organic sludge and nitrate-rich water from recirculating systems can be
avoided by subjecting the sludge to anaerobic digestion and using the digestion
products to fuel denitrifiers for removal of nitrate. Essentially, by means of such
treatment, organic matter is converted to carbon dioxide and inorganic nitrogen to
nitrogen gas. In the present study it is demonstrated that by means of this treatment
method a considerable fraction of the added phosphorus is trapped in the denitrify-


Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

123

ing bacterial biomass. Differences between bacterial phosphate accumulation by
denitrifiers in this system and in EBPR systems are discussed.

2. Materials and methods

2.1. Experimental treatment system
A small prototype treatment system (Fig. 1) was operated at our facilities at the
Rehovot campus. The system was comprised of two basins (500 l, each), a
denitrifying fluidized bed reactor and a nitrifying trickling filter. One basin served
as a digestion basin. From this basin, water was pumped at a rate of 6.0 l min − 1
into the fluidized bed reactor (height: 198 cm; diameter: 6.1 cm; volume: 5.8 l)
containing 3 l of sand (average diameter: 0.7 mm) as bacterial carrier material.
Effluent water from the fluidized bed reactor flowed into the trickling filter basin
situated underneath a nitrifying, trickling filter (volume: 1 m3) consisting of PVC
cross-flow medium with a specific surface area of 240 m2 m − 3 (Jerushalmi, Israel).
Water from this basin was pumped over the trickling filter at a rate of 42.0 l min − 1.
From the trickling filter basin, water was returned by gravity to the digestion basin.
The system was operated for a period of 210 days with a weekly water exchange of
9 10% of the total water volume. Periodically, part of the biofilm developing in the
fluidized bed reactor was harvested by removing a portion of the sand, cleaning it
of biofilm growth and returning the cleaned sand once more to the fluidized bed
reactor. Weekly, four to five daily portions of 400 g feed (30% protein; 1%
phosphorus-P) were added to the digestion basin (total number of recorded feeding
days: 144).

Fig. 1. Schematic presentation of the experimental treatment system (not to scale).


124

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

2.2. Phosphate accumulation and nitrate remo6al by denitrifying isolates

Three denitrifying strains, isolated from a fluidized bed reactor used for nitrate
removal in a recirculating fish culture system (van Rijn et al., 1995; Aboutboul et
al., 1996), were tested for combined nitrate and phosphate removal. Based on fatty
acid profiles and 16S-rDNA, two of these isolates could be identified as Pseudomonas aeruginosa and Paracoccus denitrificans. The other, a Pseudomonas isolate,
could not be identified to species level and was deposited as Pseudomonas sp. strain
JR12 (DSM c12019) in the German Collection of Microorganisms and Cell
Cultures (van Rijn et al., 1996). The denitrifying organisms were cultured at 30°C
in medium containing (per liter): Sodium acetate, 5.6 g; KH2PO4, 0.4 g; NH4Cl, 1.0
g; MgSO4.7H2O, 0.6 g; Na2S2O3 5H2O, 0.1 g; CaCl2 2H2O, 0.07 g, Tris (Hydroxymethyl aminomethane) – hydrochloride, 12 g; and 2 ml of a trace element solution
(Visniac and Santer, 1975). The pH of the medium was 7.2. Studies were conducted
with cells harvested during the late log phase of growth (after 4–5 days). Cells were
washed twice and resuspended in the aforementioned synthetic medium with
various phosphate and nitrate (as KNO3) levels (Section 3). Determinations of
nitrate and phosphate removal by these isolates under various conditions were
conducted in triplicate in a temperature-controlled (30°C) incubation vessel (300
ml), placed on a magnetic stirrer and fitted with nitrate, pH and oxygen/temperature electrodes. Anaerobic conditions in the vessel were obtained by continuous
flushing with prepurified nitrogen gas. Positive pressure within the incubation vessel
prevented oxygen penetration, as verified by continuous oxygen monitoring. The
experiments were initiated by acetate addition. Periodically, samples were withdrawn, filtered, and analyzed for ammonia, nitrite and phosphorus. Changes in
nitrate levels and pH were monitored every 2–5 min, whereas protein concentrations were determined in aliquots withdrawn at the beginning and end of the
experiment. During the various experiments, the bacterial biomass (as measured by
protein analysis) did not increase by more than 20%. Ammonia concentrations
decreased in correspondence with the increase in bacterial biomass in the medium.
An increase in pH (not exceeding 0.6 units) was measured in all experiments.

2.3. Batch studies with bacterial consortia obtained from the laboratory-scale
treatment system
Organic matter, making up the biofilms on the PVC and sand carriers in the
trickling filter and fluidized bed reactor, respectively, was detached from the carriers
by grinding. After washing the detached biofilms in the above described medium,

combined nitrate and phosphate removal by the bacterial consortia present in these
biofilms was examined by the same experimental protocol used for the bacterial
isolates.

2.4. Quantitati6e and qualitati6e phosphorus analyses
Toward the end of the experimental period (between days 200 and 210),


Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

125

triplicate samples were derived from the fluidized bed reactor (200 g colonized
sand), from the trickling filter (36 cm2 of colonized PVC), from the digestion basin
(3 ml sludge) and from the water in the treatment system (2 l). Dry weight and total
phosphorus content of the organic matter present in the samples were determined.
Total phosphorus in each of the treatment compartments was calculated based on
the following information: total sand dry weight in fluidized bed reactor, 800 g;
total surface area of PVC in trickling filter, 240 m2; total sludge dry weight in
digestion basin, 8.8 kg; total water volume in treatment system, 1000 l.

2.5. Analytical procedures
Inorganic nutrients were determined in GF/C (Whatman, UK) filtered samples.
Total ammonia (NH3 and NH+
4 ) was determined as described by Scheiner (1976),
nitrite according to Strickland and Parsons (1968) and nitrate was measured with
the Szechrome NAS reagent (Ben Gurion University, Applied Research Institute)
or, in laboratory batch experiments, with a specific nitrate electrode (Radiometer,
Denmark) amplified with a pH meter (Radiometer, model: PHM92). Inorganic
orthophosphate (phosphate throughout the text) in filtered samples and total

phosphorus (organic, particulate and inorganic orthophosphate) in unfiltered samples was determined with the ascorbic acid method described by Golterman et al.
(1978).
Oxygen and temperature were measured with a YSI (model 57) temperature/oxygen probe (Yellow Springs Instruments, USA).
Bacterial dry weight and dry weight of organic matter obtained from the various
components of the treatment system were determined after overnight drying of the
samples at 105°C. Protein was determined according to Lowry et al. (1951) with
bovine serum albumin as standard.

3. Results

3.1. Inorganic nitrogen and phosphorus concentrations in the experimental
treatment system
Despite the closed-mode of operation and daily feed supply, inorganic nitrogen
and phosphate concentrations in the experimental system (determined in samples
obtained from the trickling filter basin), did not accumulate over the 210 days of
operation (Fig. 2). Oxygen was at saturation in the trickling filter basin while it was
undetectable in the digestion basin and fluidized bed reactor (not shown). The pH
of the system fluctuated between 6.9 and 7.6 (not shown).
Due to the relatively short retention times, differences between inlet and outlet
concentrations of inorganic nitrogen and phosphate in each of the treatment
compartments (Fig. 3) were often below the analytical limit of detection. However,


126

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

Fig. 2. Ammonia, nitrite, nitrate and phosphate concentrations in the experimental treatment system
during 210 days of operation.

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

127

Fig. 3. Removal (positive values) or production (negative values) of ammonia, nitrite, nitrate and
phosphate by the fluidized bed reactor (F.B.R.), trickling filter (T.F.) and digestion basin (D.B.) over the
experimental period.

it is evident that ammonia was removed in the trickling filter and the fluidized bed
reactor and produced in the digestion basin. Nitrate was produced in the trickling
filter, removed in the fluidized bed reactor and removed or produced in the
digestion basin. The fluidized bed reactor removed phosphate while it was removed
or produced in the trickling filter and digestion basin.


128

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

A qualitative analysis revealed that samples derived from the fluidized bed
reactor were high in phosphorus content as compared to samples from other
treatment compartments (Table 1). Absolute values of total phosphorus were
highest in the trickling filter and much lower in the digestion basin, fluidized bed
reactor and in the water body of the system (Table 1). The total phosphorus load
of the system was 576 g P (144 feeding days × 4 g P). We estimated that during the
experimental period 16 g P was removed with water exchange and 49 g P with the
wasted biomass from the fluidized bed column. Therefore, the total expected
phosphorus content of the system over the experimental period was 511 g P. Based
on results presented in Table 1, this implies that as much as 9198% of the added

phosphorus was retained within the trickling filter, 0.79 0.1% in the fluidized bed
reactor, 4.79 1.9% in the digestion basin and 2.2 9 0.5% in the water of the system.

3.2. Phosphate accumulation by denitrifying and nitrifying consortia
A denitrifying consortium derived from the fluidized bed reactor was incubated
under laboratory conditions in the presence or absence of nitrate (Fig. 4). Phosphate uptake took place in the presence of nitrate whereas after depletion of nitrate
from the medium, phosphate was released. In the presence of nitrate, the consortium assimilated ammonia and phosphate at a molar N/P ratio of 1.9. Taking into
account that the molar N/P ratio of bacterial biomass varies from 5 to 16 (Brock
and Madigan, 1991), it can be concluded that in the presence of nitrate, phosphate
is assimilated in excess of the metabolic requirements of the bacteria comprising the
consortium.
A nitrifying consortium derived from the trickling filter was incubated in the
laboratory under aerobic (nitrifying) and anoxic (denitrifying) conditions. Anoxic
incubation was conducted in the presence of acetate. Under aerobic conditions (Fig.
5(A)), ammonia was nitrified to nitrate while phosphate concentrations in the
medium increased gradually. Trickling filter material was rich in organic matter.
Degradation of this organic matter and ammonification of nitrogenous organic
compounds to ammonia probably explains the deficit observed between nitrate
production and ammonia consumption. Incubation of trickling filter material in the
absence of oxygen resulted in a decrease of nitrate and phosphate concentrations in
the medium (Fig. 5(B)).
Table 1
Absolute and relative phosphorus content ( 9S.D.) in different compartments of the experimental
treatment system after 210 days of operation
Compartment
Fluidized bed reactor
Trickling filter
Digestion basin
Water body

Total phosphorus (g) Phosphorus in organic matter (mg/g dry weight)
3.59 0.3

118.5 95.3

464.7 9 40.0
24.2 99.7
11.0 92.6

18.6 9 5.2
2.4 9 1.1
9.0 9 1.5


Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

129

Fig. 4. Changes in nitrate ( ) and phosphate ( ) concentrations during batch incubation of a
denitrifying consortium derived from the fluidized bed reactor. Arrow indicates time of nitrate addition.

3.3. Phosphate accumulation by denitrifying isolates
All three isolates assimilated phosphate in excess under denitrifying conditions
(i.e. in the presence of nitrate) and released phosphate when nitrate became
depleted (Fig. 6). Molar N/P ratios of the three different isolates at the end of the
denitrifying period were 0.4, 2.1 and 1.3 for P. aeruginosa, P. denitrificans and
Pseudomonas sp. (JR12), respectively.

3.4. Phosphate accumulation and nitrate remo6al in the fluidized bed reactor
The performance of the fluidized bed reactor on selected days throughout the 210

days experimental period are presented in Table 2. It is shown that the denitrifying
consortium present in this reactor, assimilated ammonia and phosphate at a molar
N/P ratio ranging from 0.5 to 2.4; i.e. phosphate accumulation by this consortium
was in excess of the metabolic requirements. With undetectable low inlet concentrations of nitrate and nitrite (days 45 and 46), phosphate was released as indicated by
the negative phosphate removal values. The latter observation points to the fact
that only under denitrifying conditions, i.e. in the presence of nitrate, the denitrifying consortium was capable of phosphate uptake in excess of metabolic
requirements.
Finally, the significance of anoxic treatment (fluidized bed reactor and digestion
basin) with respect to phosphate removal was demonstrated by disconnecting this
stage from the aerobic (trickling filter) treatment stage (Fig. 7). Without supply of
nitrate-rich water, levels of nitrate dropped rapidly in the digestion basin. Under
the resulting anaerobic conditions, biologically-stored phosphorus was released as
can be seen from the rapid increase in phosphate concentrations in the medium.


130

Day from start of operation

NO3 (mmol m−2
day−1)

NO2 (mmol m−2
day−1)

NH4 (mmol m−2 day−1)

PO4 (mmol m−2 day−1)

NH4/PO4

43
44
45
46
79
85
86
134
148
153

274
167
0
0
333
320
108
485
392
571

−92
−64
0
0
0
135
230

−124
−14
−10

31
32
2
70
21
60
50
36
56
36

53
66
−10
−53
20
24
36
26
68
15

0.6
0.5
−0.2
−1.3

1.1
2.5
1.4
1.4
0.8
2.4

a

Negative values indicate accumulation.

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

Table 2
Daily removal rates (per surface area of carrier) of inorganic nitrogen and phosphate by the fluidized bed reactor and calculated ratios between ammonia
and phosphate removal at selected sampling daysa


Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

131

Subsequent reconnection of the water supply was followed by a nitrate increase and
a phosphate decrease in the anoxic treatment stage.

4. Discussion
Information on phosphorus dynamics in recirculating fish culture systems is
scarce. As determined in more conventional fish culture systems, phosphorus
recovery values by fish vary from 10 to 30% of the phosphorus added in the feed
(Avnimelech and Lacher, 1979; Boyd, 1985; Schroeder et al., 1991). Of the released

phosphorus, roughly 20% is in the soluble form while the remainder is present in

Fig. 5. Changes in ammonia ( ), nitrite ( ), nitrate (
) and phosphate ( ) concentrations during
aerobic (A) and anoxic (B) batch incubation of a bacterial consortium derived from the trickling filter.
Anoxic incubation was conducted in the presence of acetate.


132

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

Fig. 6. Changes in nitrite ( ), nitrate ( ) and phosphate (
) concentrations during batch incubation
of: (A) P. aeruginosa, (B) P. denitrificans and (C) Pseudomonas sp. (JR12). Arrow indicates time of
nitrate addition.


Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

133

Fig. 7. Changes in nitrate ( ) and phosphate ( ) concentrations in the digestion basin after uncoupling
(nitrate off) and reconnecting (nitrate on) the water supply from the trickling filter to the digestion basin.

the organic sludge (Bodvin et al., 1996). Chen et al. (1996) estimated that the total
phosphorus in aquaculture sludge is as high as 1.3% of the total solids. Increased
environmental concern associated with phosphorus discharge has stimulated research on phosphorus reduction in aquaculture systems. Most of the studies in this
field have dealt with decreasing phosphorus inputs by increasing the dietary
phosphorus availability (Rodehutscord et al., 1994; Rodehutscord and Pfeffer,

1995). Phosphorus released into the culture systems is generally left untreated and
discharged with the organic solids and effluent water.
In the present study it is demonstrated that crude denitrifying consortia and
denitrifying isolates are capable of phosphorus storage in excess of their metabolic
requirements. By Niesser staining and electron microscopy we found that this
excess phosphorus is stored by these bacteria as polyphosphate (Barak and van
Rijn, in press). As mentioned in the introduction section, the presence of
denitrifying poly-P organisms in EBPR processes used for wastewater treatment has
recently been demonstrated. Information on these organisms is mainly derived from
mass balance studies on changes in inorganic nitrogen and phosphorus in crude
sludge samples exposed to alternating aerobic/anoxic conditions or from similar
balances on enrichment cultures (Kuba et al., 1997). From these studies it can be
concluded that PHA-mediated polyphosphate synthesis in the aerobic or anoxic
treatment step is a common trade among all the organisms thought to be involved
in the phosphorus dynamics in these systems. Unfortunately, this evidence is
circumstantial since, so far, efforts to obtain pure, axenic cultures of either
denitrifying or non-denitrifying poly-P organisms have been unsuccessful (Mino et
al., 1998).


134

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

Based on the information available on poly-P organisms in EBPR processes we
assume that these bacteria are different from the denitrifying organisms responsible
for polyphosphate accumulation in the present study. This assumption is based on
the fact that the denitrifying organisms present in the fluidized bed reactor in this
study were shown to be capable of polyphosphate synthesis under permanent
anoxic conditions. Further evidence for this assumption was obtained from recent

studies in our laboratory. Studies on a number of denitrifying strains isolated from
the fluidized bed reactor revealed that: (a) polyphosphate synthesis and denitrification were conducted by organisms incapable of producing PHA; and (b) in
organisms capable of producing PHA, polyphosphate synthesis was not coupled to
PHA and glycogen degradation (Barak and van Rijn, in press).
A characterization of the phosphate dynamics in the experimental treatment
system revealed that, although most of the phosphorus accumulated in the trickling
filter, active removal of phosphate was highest in the fluidized bed reactor. The
phosphorus content of organic matter attached to the sand particles in the fluidized
bed reactor was as much as 11.8% of the dry weight. Similar values were reported
for polyphosphate accumulating organisms in wastewater treatment plants (Degre´mont Ltd., 1991). The high total phosphorus content of the trickling filter sludge
can be explained as follows. As our experimental set-up contained no mechanical
filtration stage, water led into the trickling filter was rich in organic matter. It may
be assumed, therefore, that much of the sloughed denitrifying biomass from the
fluidized bed reactor was captured in the trickling filter. This, together with the
anoxic areas within the trickling filter resulting from the high organic load, may
have resulted in a considerable accumulation of denitrifiers in the trickling filter.
Evidence for this assumption was provided by the observed denitrification potential
of the trickling filter material upon batch incubation under anoxic conditions (Fig.
5(B)). It is interesting to notice that aerobic incubation of the trickling filter
material resulted in a release of phosphorus into the surrounding medium (Fig.
5(A)). A possible explanation for this observation is that phosphorus was released
due to carbon limitation of the denitrifying organisms since no external carbon was
added. We obtained similar results (phosphorus release under conditions of carbon
limitation) in batch experiments with denitrifying isolates (not shown).

5. Conclusions
Based on the presented results we conclude that denitrifying organisms are
capable of phosphate uptake in excess of their metabolic requirements. It seems
feasible therefore, that this property may provide a biological means of phosphate
removal from aquaculture systems in which water treatment comprises a denitrifying stage. By harvesting part of the denitrifying biomass one might be able to

withdraw particulate phosphorus. In the treatment system examined in this study,
the trickling filter was operated under conditions far from optimal as it received
unfiltered water, rich in organic material. This together with the large size of the
trickling filter relative to the water volume of the system caused much of the


Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

135

particulate organic matter to be trapped in this filter. It remains to be examined,
therefore, how the addition of a mechanical filtration unit between the fluidized bed
reactor and the trickling filter, could serve as a trap for phosphate-rich organic
material, and if such phosphate removal will be significant in controlling the
inorganic phosphate levels in the culture and effluent water of recirculating systems.

Acknowledgements
This study was supported through grant number 820-0136-98 by the Chief
Scientist Office, Ministry of Agriculture and Rural Development, Israel.

References
Aboutboul, Y., Arbiv, R., van Rijn, J., 1996. Anaerobic treatment of fish culture effluents: volatile fatty
acid mediated denitrification. Aquaculture 133, 21 – 32.
Avnimelech, Y., Lacher, M., 1979. A tentative nutrient balance for intensive fish ponds. Bamidgeh 31,
3 – 8.
Barak, Y., van Rijn., 2000. Atypical polyphosphate accumulation by the denitrifying bacterium
Paracoccus denitrificans. Appl. Environ. Microbiol., in press.
Barker, P.S., Dold, P.L., 1996. Denitrification behaviour in biological excess phosphorus removal
activated sludge systems. Water Res. 30, 769 – 780.
Bodvin, T., Indergaard, M., Norgaard, E., Jensen, A., Skaar, A., 1996. Clean technology in aquaculture:

a production without waste products. Hydrobiologia 327, 83 – 86.
Boyd, C.E., 1985. Chemical budget for channel catfish ponds. Trans. Am. Fish. Soc. 114, 291 – 298.
Brock, T.D., Madigan, M.T., 1991. Biology of Microorganisms, 6th ed. Prentice Hall, New Jersey.
Chen, S, Stechey, D, Malone, R.F., 1994. Suspended solids control in recirculating aquaculture systems.
In: Timmons, M.B., Losordo, T.M. (Eds.), Aquaculture Water Reuse Systems: Engineering Design
and Management. Development in Aquaculture and Fisheries Sciences, vol. 27. Elsevier, Amsterdam,
pp. 61 –100.
Chen, S., Ning, Z., Malone, R.F., 1996. Aquaculture sludge treatment using anaerobic and facultative
lagoon system. In: Libey, G.S., Timmons, M.B. (Eds.), Successes and Failures in Commercial
Recirculating Aquaculture. Northeast Regional Agricultural Extension Service, Ithaca, NY, pp.
421 – 430.
Degre´mont Ltd., 1991. Water Treatment Handbook, vol. 1, 6th ed. Lavoisier, Paris.
Golterman, H.L., Clymo, R.S., Ohnstad, M.A.M., 1978. Methods for Physical and Chemical Analysis
of Fresh Waters. IBP Handbook No. 8, 2nd ed. Blackwell, Oxford.
Kuba, T., Smolders, G., van Loosdrecht, M.C.M., Heijnen, J.J., 1997. A metabolic model for biological
phosphorus removal by denitrifying organisms. Biotech. Bioeng. 52, 685 – 695.
Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurements with the
Folin-phenol reagent. J. Biol. Chem. 193, 265 – 275.
Mino, T., van Loosdrecht, M.C.M., Heijnen, J.J., 1998. Microbiology and biochemistry of the enhanced
biological phosphate removal process. Water Res. 32, 3193 – 3207.
Rodehutscord, M., Pfeffer, E., 1995. Effects of supplemental microbial phytase on phosphorus digestibility and utilization in rainbow trout (Oncorhynchus mykiss). Water Sci. Technol. 31, 143 – 147.
Rodehutscord, M., Mandel, S., Pfeffer, E., 1994. Reduced protein content and use of wheat gluten in
diets for rainbow trout: effect on water loading with N and P. J. Appl. Ichthyol. 10, 271 – 273.
Scheiner, D., 1976. Determinations of ammonia and Kjeldahl nitrogen by the indophenol method.
Water Res. 10, 31–36.


136

Y. Barak, J. 6an Rijn / Aquacultural Engineering 22 (2000) 121–136

Schroeder, G.L., Alkon, A., Laher, M., 1991. Nutrient flow in pond aquaculture systems. In: Brune,
D.E., Tomasso, J.R. (Eds.), Advances in World Aquaculture, Volume 3, Aquaculture and Water
Quality. The World Aquaculture Society, Baton Rouge, LA, pp. 489 – 505.
Strickland, J.D., Parsons, T.R., 1968. A practical handbook of seawater analysis. Fish. Res. Board
Canada 167, 77–80.
Toerien, D.F., Gerber, A., Lotter, L.H., Cloete, T.E., 1990. Enhanced biological phosphorus removal in
activated sludge systems. In: Marshall, K.C. (Ed.), Advances in Microbial Ecology, vol. 11. Plenum,
New York, pp. 173–219.
van Rijn, J., Fonarev, N., Berkowitz, B., 1995. Anaerobic treatment of fish culture effluents: digestion
of fish feed and release of volatile fatty acids. Aquaculture 133, 9 – 20.
van Rijn, J., Tal, Y., Barak, Y., 1996. Influence of volatile fatty acids on nitrite accumulation by a
Pseudomonas stutzeri strain isolated from a denitrifying fluidized bed reactor. Appl. Environ.
Microbiol. 62 (7), 2615–2620.
Visniac, W., Santer, M., 1975. The Thiobacilli. Bacteriol. Rev. 21, 195 – 213.

.

.