Bioresource Technology 101 (2010) 1511–1517

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

A study on the optimal hydraulic loading rate and plant ratios in recirculation
aquaponic system
Azizah Endut a,*, A. Jusoh b, N. Ali b, W.B. Wan Nik c, A. Hassan d
a
Faculty of Innovative Design and Technology, University Darul Iman Malaysia, KUSZA Campus, 21300 Kuala Terengganu, Malaysia
b
Faculty of Science and Technology, University Malaysia Terengganu, 21030 Kuala Terengganu, Malaysia
c
Faculty of Maritime Studies and Marine Science, University Malaysia Terengganu, 21030 Kuala Terengganu, Malaysia
d
Institute of Tropical Aquaculture, University Malaysia Terengganu, 21030 Kuala Terengganu, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: The growths of the African catfish (Clarias gariepinus) and water spinach (Ipomoea aquatica) were evalu-
Received 4 June 2009 ated in recirculation aquaponic system (RAS). Fish production performance, plant growth and nutrient
Received in revised form 10 September removal were measured and their dependence on hydraulic loading rate (HLR) was assessed. Fish produc-
2009
tion did not differ significantly between hydraulic loading rates. In contrast to the fish production, the
Accepted 10 September 2009
Available online 9 October 2009
water spinach yield was significantly higher in the lower hydraulic loading rate. Fish production, plant
growth and percentage nutrient removal were highest at hydraulic loading rate of 1.28 m/day. The ratio
of fish to plant production has been calculated to balance nutrient generation from fish with nutrient
Keywords:
Hydraulic loading rate
removal by plants and the optimum ratio was 15–42 gram of fish feed/m2 of plant growing area. Each
Aquaponic unit in RAS was evaluated in terms of oxygen demand. Using specified feeding regime, mass balance
African catfish equations were applied to quantify the waste discharges from rearing tanks and treatment units. The
Water spinach waste discharged was found to be strongly dependent on hydraulic loading rate.
Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction production. The amount of nitrate produced in a fish culture sys-

tem is directly proportional to two factors: the amount or den-
Aquaculture probably the fastest growing food-producing sec- sity of fish in the system and the amount and protein content
tor, now accounts for almost 50% of the world’s food fish and is of the food, as different fish species require different protein
perceived as having the greatest potential to meet the growing de- content in their respective diets.
mand for aquatic food. It is estimated that at least an additional 40 Integrated systems use water more efficiently through the
million tonnes of aquatic food will be required by 2030 to maintain interacting activities of fish and plants. The addition of water to a
the current per capita consumption (FAO, 2006). fish tank to satisfy the oxygen requirements depends on the oxy-
When fish are cultured, only a small proportion of the feed is gen consumption of the fish, the oxygen concentration in the inlet
converted (25–30%) to useable energy (Rakocy et al., 1993). The water and the lowest acceptable concentration in the outlet water
balance of nutrients is excreted in solid and dissolved fractions. (Lekang, 2007). Hence effective HLR can be employed to achieve
Dissolved nutrients accumulate in recirculation systems with low optimal growth for the fish and plants.
water exchange and high feeding rates to levels which approxi- The rate of change in nutrient concentration can be influenced
mate hydroponic nutrient solutions. by varying the ratio of plants to fish (Rakocy et al., 2006). However,
Recirculation aquaponic system (RAS) is a promising technol- since the relative proportions of soluble nutrients made available
ogy in the integration of fish and hydroponic plant production. to the hydroponic plants by fish excretion do not mirror the pro-
The fish water, rich in nutrients is used for plant growth, while portions of nutrients assimilated by normally growing plants, the
the plants are used as biofilters for water regeneration. Whilst rates of change in concentration for individual nutrients differ.
biofiltration converts the harmful into the harmless, the end The disparity in accumulation or reduction rates of different nutri-
point is a buildup of nutrients within recirculation systems, prin- ents quickly results in suboptimal concentrations and ratios of
cipally consisting of nitrates and phosphates. Nutrient removal nutrients, thereby reducing the nutritional adequacy of the solu-
by plants improves the quality of effluent and may enhance fish tion for plants. Theoretically, the nutrient content of a diet can
be manipulated to make the relative proportions of nutrients ex-
* Corresponding author. Tel.: +609 6653519; fax: +609 6673523. creted by fish more similar to the relative proportions of nutrients
E-mail address: [email protected] (A. Endut). assimilated by plants. With such a diet, there would be an optimal

0960-8524/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2009.09.040
1512 A. Endut et al. / Bioresource Technology 101 (2010) 1511–1517

ration of fish to plants and optimal nutrient supplementation (Sea- and pH were in the range of 27.5–28.8 °C and 5.6–7.3 respectively
wright et al., 1998). and was acceptable for African catfish.
Several mass balance models have been proposed from previous Samples of C. gariepinus fingerlings, with an initial average body
studies (Pagand et al., 2000; Papatryphon et al., 2005; Schneider weight of 30–40 g were randomly transferred into three replicate
et al., 2005; Mongirdas and Kusta, 2006), from which the total fiberglass culture tanks. Water exited and flowed from the culture
nitrogen and phosphorus discharges into receiving waters can be tank was sprinkled over the vegetables in the hydroponic trough
estimated. However, most of these studies were conducted in open and outflow trickled down to the sump for denitrification process.
systems. Recently, the incorporation of recirculated fish with veg- The components were installed such that the water flowed by
etable hydroponics production has become an interesting model to gravity, by placing components at appropriate elevation relative
private sector, aquaculture and environmental scientists (Rakocy to one another. The water was then pumped vertically to the sand
et al., 2006; Bakhsh and Shariff, 2007; Endut et al., 2009). filtration tanks for particulate removal. After exiting the sand filter
The objectives of this study were to (1) determine the optimum the water went directly to water storage tank and was flowed by
hydraulic loading rate in term of fish production, plant production, gravity back to the fish tank. During this study the inflow rates
and nutrient removal, (2) evaluate the optimum plants ratio in of the RAS were maintained to be identical as possible by adjusting
term of daily fish feed input to plant growing area, and (3) study the gate valves according to the target rate of each trial.
the mass balance of oxygen in achieving sustainable balance be- In the first experiment, five trials were consecutively performed
tween fish and plants. at different hydraulic loading rates, each operating for 35 days and
compared with control with no plants. Each treatment was repli-
cated thrice. The experimental was operated with fish for one week
2. Methods prior to the initiation in order to acclimate the biofilters as to min-
imize net nutrient uptake by bacteria at the beginning of each trial.
2.1. Experimental design At the initiation of hydraulic loading rate of 0.64 m/day, system
was flushed and fingerlings African catfish were added to each cul-
The recirculation aquaponic system (RAS) utilized is depicted in ture tank up to the treatment biomass. Due to our inability to
Fig. 1. The experimental facility was located in a greenhouse of the obtain fish of similar size for trials two and three, fish from the
University of Malaysia Terengganu campus. RAS consisted of a prior trial were pooled and reallocated to the systems for the fol-
fiberglass rearing tank, hydroponic trough (growing bed), sand fil- lowing trial. Hand feeding, twice per day at the range of 2–4% body
ter for solid removal, sump system for denitrification unit, water weight/day. Fish were fed with 3.2 mm commercial diet floating
holding tank and reservoir (pre-aeration). Pipelines made of poly- pellet (Cargill Company) with 32% protein and 10% moisture. The
vinyl chloride were installed to connect the culture tank and food size was adjusted to compensate for changes in fish size.
hydroponic trough to recirculate the water. Water spinach seedlings were planted directly into the gravel sub-
Three culture tanks arranged in series were used in the rearing strate of the hydroponic growing bed at 10 cm  10 cm spacing.
of African catfish (Clarias gariepinus). Air stones, connected to an air In the second experiment, the effect of seven ratios of plants to
blower were installed in the culture tank to supply oxygen for fish fish (2, 4, 6, 7, 8, 9 and 10) was evaluated by manipulating water
culture. Water level in each culture tank was kept at 0.85 m deep spinach stocking rate. The size of the hydroponic trough and sys-
to maintain the water volume at 1000 L. Water lost through evap- tem volume were the same for all ratios, but the daily feeding rate
oration, transpiration and sludge removal was replenished with increased in direct proportion to the fish biomass. African catfish
water in the pre-aeration tank. The tank openings were covered were stocked into rearing tank as at optimum HLR obtained in
by plastic net (20 mm aperture) to hinder the fish jumping out of the first experiment. Water samples from outlet of fish tank and
the tanks. Measurements of temperature, dissolved oxygen (DO) outlet of the hydroponic trough were collected once a week for
and pH of water samples were performed in situ during the sam- water quality determination. At the end of the experiment, all
pling process using the YSI multi-probe meter (model YSI 550A) fishes in each tank will be netted, weighed and their individual
and pH Cyber Scan waterproof, respectively. Water temperature lengths will be recorded as well as the weight of vegetable.

Fig. 1. A: Culture tank, B: hydroponic trough (planted bed), C: hydroponic trough (control bed), D: filter, E: sprinkler, F: sump, G: pump, H: rapid sand filter, I: water storage
tank, J: air blower, and K: valves.
 A. Endut et al. / Bioresource Technology 101 (2010) 1511–1517 1513

If nutrient flows are subsequently connected from one system to 2.3. Samples analyses
another one, overall system nutrient retention and balances can be
estimated. The oxygen flows and its retentions were calculated using Water samples were taken once a week from each culture tank,
mass balances, based on the concept: output = input  retention. influent and effluent of the hydroponic trough, sump, water stor-
This retention can be expresses as g per kg feed with the feed to age tank and inflow of culture tank. The samples were analyzed
the fish (percent fish nutrient). The nutrient discharges (output) for 5-day biochemical oxygen demand (BOD5), total suspended so-
from the system serve as input in the other subsequent system. Table lid (TSS), total ammonium nitrogen (TAN), nitrite nitrogen
1 illustrates the hydraulic condition for this study. (NO 
2 —N), nitrate nitrogen (NO3 —N) and total phosphorus (TP).
Dissolved oxygen, pH, and temperature were also recorded each
time water was collected. Weekly sampling was carried out be-
2.2. Measurement of growth and yield
tween 8.30 am and 9.30 am in each sampling date and refrigerated
at 4 °C in labeled polythene bottles for chemical analysis. The BOD5
To assess the overall system performance, data on fish growth
and TSS analyses were performed according to Standard Method
and feeding were collected. The feeding data collected include,
(APHA 1998). The TAN, NO 
2 —N, NO3 —N and TP measurements
the feeding rate, amount of feed; number of feeding/day; feed
were performed using HACH DR4000 spectrophotometer accord-
amount per tank per day; total feed per day and feed protein. Fish
ing to salicylate, diazotization, cadmium reduction and ascorbic
(10%) were taken from the culture tank to measure their length and
acid method respectively. The DO and pH of the sample were mea-
body weight to estimate the growth rate of the fish and assumed to
sured using DO meter YSI 55A and pH cyber scan waterproof
be representative of the fish in the tanks. The growth of fish was
respectively.
also monitored from the time at stocking up till the harvest time.
Fish sampling in juvenile and grow-out systems were done on a
weekly basis for survival and average weight. 2.4. Statistical methods
The following production parameters were determined accord-
ing to the procedure of Ridha and Cruz (2001): Statistical software of Statistical Package for the Social Sciences
(SPSS) Version 16 and Microsoft Excel were used to calculate mean,
ðln mean final weight  ln mean initial weightÞ  100 standard deviation, and one-way ANOVA. Differences of mean
Specific growth rateðSGRÞ ¼
culture days were evaluated for significance by the range tests of Tukey HSD
(p 6 0.05) for homogeneous variances (Levene test) and by the
range test of Dunnett T3 (p 6 0.05) for inhomogeneous variances,
respectively (Schulz et al., 2003).
total weight of dry feed given
Feed conversion ratio ðFCRÞ ¼
total wet weight gain
3. Results and discussion
Plants growth was monitored weekly by measuring the plant
height of all water spinach plants on a 0.5 m2 planting area accom- 3.1. Effect of hydraulic loading rates
panied by counting of number of shoots. The plants were harvested
at height ranging from 45 to 50 cm. Each growing trough was Specific growth rates (SGRs), feed conversion ratio (FCR) and
cleaned and the biomass of plants was measured and recorded. fish production did not differ significantly between hydraulic load-
ing rates (Table 2). FCR values are in the range of 1.23–1.39. In our
study, the same feed is used and the ration is fixed similarly in all
Table 1
Hydraulic condition for operating the aquaponic recirculation system. culture tanks. Stocking at hydraulic loading rate of 1.28 m/day
gives the best production performance (Table 2).
Q (m3 day1) HLRa (m day1) HRTb for overall system (h)
The FCR recorded (1.23–1.39) is not far above the ideal value of
4.6 0.64 4.5 1.0 for culture of African catfish in recirculation system and FCR va-
9.2 1.28 2.3
lue 0.85 reported in the culture of African catfish by Eding and
13.8 1.92 1.5
18.4 2.56 1.1 Kamstra (2001). However the recorded FCR are better than the
23.0 3.20 0.9 range 1.1–1.7 reported in recirculation system of African catfish
a
as reported by Akinwole and Faturoti (2007). HLR did not affect
HLR, hydraulic loading rate, which is flow rate (Q) divided by total surface area
of the trough.
growth rate or feed conversion ratio.
b
HRT, hydraulic retention time, which can be computed as (surface area  water Plants grew actively in the hydroponic trough and did not
depth  porosity of gravel trough/flow rate). identify any nutritional deficiencies or mineral imbalances. Plant

Table 2
Fish growth, feed conversion factor and water spinach by hydraulic loading rates.

HLRa (m/day) Fish Plant (water spinach)

SGRb (%) FCRc Production (kg/m3) Growth rated (cm/day) Production (kg/trough/harvest)
0.64 1.801 1.271 45.21 1.751 17.632
1.28 1.831 1.231 45.71 2.503 17.902
1.92 1.731 1.331 44.31 2.061 17.532
2.56 1.731 1.341 44.11 1.902 17.031
3.20 1.681 1.391 43.31 1.902 16.831

Values given are mean from triplicate data (n = 3).

Mean with the different superscript is significantly different at the p 6 0.05 level.
a
Hydraulic loading rate calculated as HLR = Q (flow rate) divided by AW (surface area) of hydroponic trough.
b
Specific growth rate calculated as SGR = ln final weight (g)  ln initial weight (g)  100 days1.
c
Feed conversion ratio is calculated as FCR = total weight of dry feed give/total wet weight gain.
d
Growth rate = height of plant divided by day.
1514 A. Endut et al. / Bioresource Technology 101 (2010) 1511–1517

Table 3
Mean values for various parameters of water quality by the RAS.

HLR (m/day) Water quality parameter

BOD5 TSS TAN NO2–N NO3–N TP
0.64 Influent (mg/L) 6.7 74.6 12.02 0.58 19.8 17.0
Effluent (mg/L) 1.7 23 2.68 0.19 5.8 6.7
Percent removal (%) 47.31 671 64.11 67.21 62.44 50.04
1.28 Influent (mg/L) 6.7 74.4 12.04 0.56 20 17.1
Effluent (mg) 1.3 21.1 2.23 0.14 5.4 6.3
Percent removal (%) 54.52 69.52 68.41 752 64.94 52.85
1.92 Influent (mg/L) 6.8 74.8 12.01 0.56 19.9 16.9
Effluent (mg/L) 1.3 19.2 1.94 0.11 6.2 7.0
Percent removal (%) 55.42 72.32 712 80.42 60.42 47.81
2.56 Influent (mg/L) 6.9 74.4 11.99 0.57 20 17.0
Effluent (mg/L) 1.0 14.2 1.68 0.09 6.6 7.1
Percent removal (%) 61.42 793 73.32 84.22 58.53 47.53
3.20 Influent (mg/L) 6.7 73.9 11.98 0.57 20.1 17.1
Effluent (mg/L) 0.7 11.2 1.14 0.06 9.7 7.9
Percent removal (%) 65.53 82.94 78.33 89.52 42.31 42.82

Different superscript numbers within one column denote statistically significant differences (p 6 0.05).

production increased as the hydraulic loading rate increased from 2008). The optimum hydraulic loading rate can be determined by
0.64 m/day to 1.28 m/day, whereas an increase in the HLR from a compromise between fish and plant productions and removal
1.28 m/day to 3.20 m/day did not result in a higher plant produc- efficiency.
tion. At the end of the growth period (20–28 days), the plants Similar to previous studies (Cottingham et al., 1999; Jamieson
reached the market size at average height of 45–50 cm. Whole et al., 2003), the improvement in TAN removal is paralleled by
plant water spinach growth rate and yields differ significantly be- the increase in NO3 —N. It can be concluded that the improvement
tween hydraulic loading rates (Table 2). in ammonia removal is due to increased nitrification activity. The
Plant growth rate and productions differ significantly between accumulation of NO 3 —N in the system indicates that after NH3–
HLR. The growth decreased significantly with increasing in HLR N is nitrified, subsequent denitrification is limited. Possible factors
supported the development of aerobic conditions in the hydro- that could limit denitrification include inadequate residence/reten-
ponic trough and hindered denitrification processes. Nevertheless, tion time for the sump to denitrify NO 3 —N, the presence of DO, or
low HLR with lower out flowing oxygen contents promoted deni- lack of available carbon in the system.
trification and highest NO 3 —N elimination is observed in lower A number of mechanisms are responsible for the removal of
hydraulic loading rate (0.64 m/day and 1.28 m/day). NO3 —N from the wastewater. One mechanism for the removal of
Average plant productions are 17.63 kg, 17.90 kg, 17.53 kg, NO3 —N is plant uptake through the root system from the growth
17.03 kg and 16.83 kg for HLR 0.64 m/day, 1.28 m/day, 1.92 m/ medium. A second mechanism for the removal of dissolved solids
day, 2.56 m/day and 3.20 m/day, respectively. The decrease in pro- is microbial assimilation. It may also be assimilated by microor-
duction corresponded strongly concludes that insufficient N in the ganisms in the water column or by biofilms associated with the
influent could be a limiting factor for a further increase in plant root mats of plants (Vaillant et al., 2004).
production. An increasing of HLR might diminish the contact time Denitrification activity is reduced if available carbon supplies
for nitrate and denitrifying bacteria, thus decreasing the perfor- were low and proceeds only when the oxygen supply was inade-
mance of hydroponic trough for denitrification (Endut et al., 2009). quate for microbial demand (Hamlin et al., 2008). In this study, car-
Snow and Ghaly (2008) evaluated the use of barley for the puri- bon availability may have been inadequate to support high levels
fication of aquaculture wastewater in a hydroponic system and re- of denitrification due to the lack of an established litter layer in
ported the crop yield was significantly influenced by the seed the system. If, on the other hand, the influent wastewater itself is
quantity. The major growth-limiting mineral is usually nitrogen an adequate source of carbon, the lack of denitrification may be
and highest growth rates and yields are generally seen when nitro- attributed to the short hydraulic retention time of the system.
gen is supplied as combination of ammonium and nitrate.
Continuous flow operation of the aquaponic system was initi-
ated with a low HLR of 0.64 m/day. The mean value and percentage 3.2. Plant ratios
removal of water quality variables at various HLR are shown in Ta-
ble 3. It is found that removal percentage of BOD5, TSS, TAN and Ni- The percentage removal values of TAN, nitrite–N, nitrate–N, to-
trite–N increased with increasing in HLR. In contrast to BOD5, TSS, tal phosphorus and plant productions at seven ratios of plants to
nitrite–N and TAN, removal percentage of nitrate–N and TP in- fish are shown in Table 4. There was significant difference in
creased with increasing in HLR from 0.64 m/day to 1.28 m/day TAN, nitrite–N, nitrate–N and total phosphorus concentrations be-
and decreased with increasing in HLR from 1.28 m/day to 3.2 m/ tween ratio of plants to fish. The percentage removal of water qual-
day. Statistically, there were significant differences in all water ity parameters and plant productions increased with increasing in
quality parameters by HLR (p < 0.05) as shown in Table 3. The plants ratios to fish until the maximum was reached at plant to fish
whole treatment, RAS basically showed effective nutrient removal ratio of 8, which was equivalent to a fish feeding rate of 15–42 g/
with average reduction efficiency range from 47% to 89.5%. m2 plant growing area. Further increasing plant ratios led to con-
Values of TSS, BOD5, TAN, nitrite–N, nitrate–N and total phos- siderable decreases in water spinach production. This strongly con-
phorus in final effluent from this study are in accordance with cludes that insufficient nitrogen in the influent of hydroponic
the previous studies (Eding and Kamstra, 2001; Schulz et al., trough could be a limiting factor for a further increase in plant
2003; Franco-Nava et al., 2004; Lin et al., 2005; Snow and Ghaly, production.
 A. Endut et al. / Bioresource Technology 101 (2010) 1511–1517 1515

Table 4
Percentage removal of water quality parameters and plant productions in RAS with hydroponic water spinach at seven ratios of plants to fish.

Fish to plants ratio (1:x) Percentage removal (%) Water spinach

x TAN NO
2 —N NO
3 —N TP Production (g WW/m2)
1 1 1 1
2 67.26 69.00 39.00 27.45 3601
4 69.932 76.602 43.322 33.482 7002
6 72.962 82.743 44.583 41.553 10003
7 79.323 90.574 51.574 48.254 11204
8 81.673 95.795 59.025 48.995 11605
9 85.864 96.535 60.205 49.015 11004
10 86.224 96.845 60.555 49.744 10003

WW, wet weight.

Values given are mean from triplicate data (n = 3); treatment with the different superscript within the same column is significantly different at the p 6 0.05 level.

In the field experiment conducted by Rakocy et al. (2006), a ra-

tio in the range of 60–100 g of fish feed/m2 of plant growing area
was used for the production of tilapia, lettuce, basil and several
other plants in raft aquaponic system. From our results we con-
clude that the technical demand on management, especially the
fish and plant species used plays a vital role in the establishment
of the configuration and relative size of integrated system compo-
nents. Plant roots, hydroponic structures and media improve water
quality by capturing solids and providing surface area for
biofiltration.

3.3. Removal rate constant

Pollutant removal can be described using first-order kinetic Fig. 2. The relationship between first-order removal rate constant for TAN (kTAN)
model (IWA, 2000). Average first-order removal rate constants and HLR.
for a specific pollutant are determined by substituting mean
hydraulic retention time (Table 1) and mean influent–effluent con-
ing rate. Efficient removal was always achieved in these studies
centrations of the pollutant (Table 3) into the following equation,
under a wide range of hydraulic loading rate because of low pollu-
and then solving for k.
tant levels of aquaculture wastewater, thus leading to HLR control-
 
Ce kehw ling the removal rate constant.
¼ expðktÞ ¼ exp ð1Þ
Ci HLR
where Ce, effluent pollutant concentration (mg/L); Ci, influent pollu- 3.4. Oxygen concentration dynamics in RAS components
tant concentration (mg/L); k, first-order removal rate constant
(day1), t, hydraulic retention time (day); HLR, hydraulic loading Figs. 3–5 show oxygen concentration dynamics in culture tank,
rate (m/day); e, porosity of hydroponic trough (assuming 0.45– influent planted trough and effluent planted trough, respectively.
0.70); and hw, water depth of trough (m). At the beginning of the system operation, the oxygen difference
These results are shown in Table 5. Distinct values of removal across the system components was insignificant because of the
rate constants for major pollutants have been reported and were low system loading (low fish biomass and therefore low feed load-
evaluated using the same methods as this study with the influ- ing). At the end of the experiment, oxygen concentration in culture
ent–effluent data. The effect of hydraulic loading rate on removal tank, influent planted trough and effluent planted trough then
rate constant was further examined by linear regression with log- reached a value of 4.70 mg/L, 4.35 and 4.2 mg/L, respectively. The
arithmic scale using the k-HLR data in Table 5. Good correlations decreased of oxygen concentration with increasing in culture day
with power function were found between removal rate constant due to the increase of fish biomass and accumulation of organic
and HLR for TAN as depicted in Fig. 2. matter in the system.
Removal rate constants for TSS, TAN, NO2–N and NO3–N, ob-
tained from this study and other comparative studies (Schulz 3.5. Oxygen consumption in fish tank
et al., 2003; Lin et al., 2005), are found to be proportional to
hydraulic loading rate to a power equation. These suggest that re- Reared fish are the main oxygen users here. Oxygen demand de-
moval rate constant would be varied depending on hydraulic load- pends on metabolic rate so the oxygen usage is expressed in terms

Table 5
First-order removal rate constants at various HLR.

HLR (m/day) Removal rate constant

kTAN (day1) kTSS (day1) kNO 1
3 —N (day ) kNO 1
2 —N (day )

0.64 4.45 5.94 4.096 5.91

1.28 7.63 12.77 8.704 10.53
1.92 11.54 17.64 12.288 13.35
2.56 11.99 24.02 16.64 21.59
3.2 12.12 27.41 17.6 30.24
1516Oxygen concentration, (mg/l) A. Endut et al. / Bioresource Technology 101 (2010) 1511–1517

10.00
HLR 0.64 (m/day)
9.00 HLR 1.28 (m/day)
HLR 1.92 (m/day)
8.00
HLR 2.56 (m/day)

7.00 HLR 3.2 (m/day)

6.00

5.00

4.00

3.00
0 5 10 15 20 25 30 35 40
Time (days)

Fig. 3. Dynamics of oxygen concentration in culture tank with different hydraulic

loading rates.
Fig. 6. Calculated and measured oxygen consumption in fish tank.

9.00
Oxygen concentration (mg/l)

HLR 0.64 (m/day)

where Q, recycled water (m3/day); Qp, supplemental water (m3/
HLR 1.28 (m/day)
8.00 day); DC, oxygen concentration differential (g/m3); ko/p, oxygen
HLR 1.92 (m/day)
used (g/kg fodder/day); W, daily feeding rate, percent from body
7.00 HLR 2.56 (m/day)
mass; m, mass of reared fish (kg), RBODT, oxygen demand by hetero-
HLR 3.2 (m/day)
trophic organisms (g/m3/day); RNT, oxygen demand of the auto-
6.00
trophic (nitrifying) microorganisms for ammonia oxidizing (g/m3/
5.00 day); WE, water exchange in the system, and 24 is the dimension
uniformly constant (day/hour). Oxygen usage dynamics is shown
4.00 in Fig. 6.
The measured mean daily usage of oxygen in the units is ex-
3.00 pressed as MO2c = 24QDC (kg/day); MO2c is the calculated oxygen de-
0 5 10 15 20 25 30 35 40 mand in the rearing units according to expression (2). The fish
Time (days) growing curve (Fig. 6) indicates the amount of the fodder fed and,
therefore, the oxygen used. Water flow should be increased with
Fig. 4. Dynamics of oxygen concentration in influent planted trough with different
the growth of fish. However, in this experiment water flow is kept
hydraulic loading rates.
the same from beginning to the end.

10.00 4. Conclusion
HLR 0.64 (m/day)
Oxygen concentration, (mg/l)

9.00 HLR 1.28 (m/day) This study demonstrated that the changes in concentrations of
HLR 1.92 (m/day) different nutrients in aquaponic system differ because of the dispar-
8.00 ity between the relative proportions of available nutrients generated
HLR 2.56 (m/day)
by fish and nutrients uptake by plants. The optimal HLR in term of
7.00 HLR 3.2 (m/day)
fish productions, plant growth and percentage nutrient removal
6.00 were found to be 1.28 m/day. Reviewing the calculated balances
and limitations of intensive integrated aquaculture systems, the per-
5.00 spectives of such integration are very promising, as these systems re-
quire fewer nutrients in relation to overall production, and reduce
4.00 nutrient discharge by reutilization. Development of the optimum
conditions, e.g., HLR, plants to fish ratio, oxygen levels and water
3.00
temperature, for system design and operation in RAS is vital in order
0 5 10 15 20 25 30 35 40
to maximize fish and plant productions and nutrient recovery and
Time (days) minimize water exchange and nutrient accumulation as well as ben-
Fig. 5. Dynamics of oxygen concentration in effluent planted trough with different
eficial environmental impacts.
hydraulic loading rates.
Acknowledgements

of feeding rate. Literature data (Bovendeur, 1989; Timmons and The authors would like to thank the Ministry of Higher Educa-
Losordo, 1994) report broad limits ranging from 200 g/kg to tion of Malaysia for financially supporting this research under bud-
610 g/kg of fodder used. Henceforth the value ko/p = 368 g O2/kg get code T-E-210-59020.
fodder/day (Bovendeur, 1989) applied for African catfish rearing
in RAS. One more portion of oxygen is required to oxidize the ex- References
cess fodder and fish excretions. Thus, according to Mongirdas
Akinwole, A.O., Faturoti, E.O., 2007. Biological performance of African Catfish
and Kusta (2006), the overall oxygen mass balance can be ex-
(Clarias gariepinus) cultured in recirculating system in Ibadan. Aquacultural
pressed as follows: Engineering 35, 18–23.
Bakhsh, Khoda, Shariff, Rashid, 2007. Nutrient optimization and computerized
24ðQ þ QpÞDC ¼ ðko=p Wm þ RBODT þ RNT Þð1  WEÞ; ð2Þ decision support program in recirculating integrated aquaculture system. The
 A. Endut et al. / Bioresource Technology 101 (2010) 1511–1517 1517

International Journal of Environmental, Cultural, Economic and Social Pagand, P., Plancheton, J.P., Casellas, C., 2000. A model for predicting the quantities
Sustainability 3 (5), 25–34. of dissolved inorganic nitrogen released in effluents from a sea bass
Bovendeur, J., 1989. Fixed-biofilm reactors applied to waste water treatment and (Dicentrarchus Labrax) recalculating water system. Aquaculture Engineering
aquaculture water recirculation systems. WAU Dissertation No. 1317, p. 171. 22, 137–153.
Cottingham, P.D., Davies, T.H., Hart, B.T., 1999. Aeration to promote nitrification in Papatryphon, E., Petit, J.S., Van der Werf, H.M.G., Kaushik, J., Kanyarushoki, C., 2005.
constructed wetlands. Environmental Technology 20, 69–75. Nutrient balance modeling as a tool for environmental management in
Eding, E., Kamstra, A., 2001. Design and performance of recirculation systems for aquaculture: the case of trout farming in France. Environmental Management
European eel and African Catfish. In: Proceeding of AES Workshop, January 23, 35 (2), 161–174.
Orlando, Florida, USA, pp. 18–28. Rakocy, J.E., Hargreaves, J.A., Bailey, D.S., 1993. Nutrient accumulation in a
Endut, A., Jusoh, A., Ali, N., Wan Nik, W.B., Hassan, A., 2009. Effect of flow rate on recirculation aquaculture system integrated with vegetable hydroponics. In:
water quality parameters and plant growth of water spinach (Ipomoea aquatica) Wang, J.K. (Ed.), Techniques for Modern Aquaculture. Proceedings of a
in an aquaponic recirculating system. Desalination and Water Treatment 5, 19– Conference, 21–23 June 1993, Spokane, WA, pp. 148–158.
28. Rakocy, J.E., Masser, M.P., Losordo, T.M., 2006. Recirculating aquaculture tank
Food and Agriculture Organization of the United Nation, 2006. The State of World production systems: aquaponics – integrating fish and plant culture. SRAC
Aquaculture. FAO Fisheries Department, Rome. p. 145. Publication No. 464.
Franco-Nava, M.A., Blancheton, J.P., Deviller, G., Charrier, A., Le-Gall, J.Y., 2004. Effect Ridha, M.T., Cruz, E.M., 2001. Effect of biofilter media on water quality and
of fish size and hydraulic regime on particulate organic matter dynamics in a biological performance of the Nile Tilapia Oreochromis niloticus L. reared in a
recirculation aquaculture system: elemental carbon and nitrogen approach. simple recirculation system. Aquacultural Engineering 24, 57–166.
Aquaculture 229, 179–198. Schneider, O., Sereti, V., Eding, E.H., Verreth, J.A.J., 2005. Analysis of nutrient flows in
Hamlin, H.J., Michaels, J.T., Beaulaton, C.M., Graham, W.F., Dutt, W., Steinbach, P., integrated intensive aquaculture systems. Aquacultural Engineering 32, 379–
Losordo, T.M., Schrader, K.K., Main, K.L., 2008. Comparing denitrification rates 401.
and carbon sources in commercial scale upflow denitrification biological filters Schulz, C., Gelbrecht, J., Rennert, B., 2003. Treatment of rainbow trout farm effluents
in aquaculture. Aquacultural Engineering 38, 79–92. in constructed wetland with emergent plants and subsurface horizontal water
International Water Association, 2000. Constructed Wetlands for Pollution Control. flow. Aquaculture 217, 207–221.
Processes, Performance, Design and Operation. IWA Publishing, London. Seawright, D.E., Stickney, R.R., Walker, R.B., 1998. Nutrient dynamics in integrated
Jamieson, T.S., Stratton, G.W., Gordon, R., Madani, A., 2003. The use of aeration to aquaculture-hydroponics systems. Aquaculture 160, 215–237.
enhance ammonia nitrogen removal in constructed wetlands. Canadian Snow, A.M., Ghaly, A.E., 2008. Use of barley for the purification of aquaculture
Biosystems Engineering 45, 1.9–1.14. wastewater in a hydroponic system. American Journal of Environmental
Lekang, O.I., 2007. Aquaculture Engineering. Blackwell Publishing Ltd., Oxford, UK. Sciences 4 (2), 89–102.
p. 352. Timmons, M.B., Losordo, T.M., 1994. Aquaculture water reuse systems: engineering
Lin, Y.F., Jing, S.R., Lee, D.Y., Chang, Y.F., Chen, Y.M., Shih, K.C., 2005. Performance of a design and management. Elsevier. p. 333.
constructed wetland treating intensive shrimp aquaculture wastewater under Vaillant, N., Monnet, F., Sallanon, H., Coudret, A., Hitmi, A., 2004. Use of commercial
high hydraulic loading rate. Environmental Pollution 134, 411–421. plant species in a hydroponic system to treat domestic wastewaters. Journal of
Mongirdas, V., Kusta, A., 2006. Oxygen mass balance in a recirculation aquaculture Environmental Quality 33, 695–702.
system for raising European Wels (Silurus glanis L.). Ekologija 4, 58–64.