International Biodeterioration & Biodegradation 85 (2013) 693e700

Contents lists available at SciVerse ScienceDirect

International Biodeterioration & Biodegradation
journal homepage: www.elsevier.com/locate/ibiod

Effects of feeding frequency and photoperiod on water quality and
crop production in a tilapiaewater spinach raft aquaponics system
Jung-Yuan Liang a, Yew-Hu Chien a, b, *
a
b

Department of Aquaculture, National Taiwan Ocean University, Keelung, Taiwan
Center of Marine Biotechnology and Bioenvironment, National Taiwan Ocean University, Keelung, Taiwan

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 25 December 2012
Received in revised form
30 March 2013
Accepted 31 March 2013
Available online 2 May 2013

A factorial arrangement of 6 treatments, 2 photoperiods for water spinach Ipomoea aquatica (12-h or
24-h light per day) X 3 feeding frequencies for red tilapia Oreochromis sp. (an equal daily ration evenly
fed 6, 4 or 2 times at 4-h, 6-h or 12-h interval, respectively) were assigned to 12 tanks. Each tank was an
aquaponics system containing fish and raft-supporting plant. Water loss in 4 weeks was 3.3%, due to leaf

transpiration mainly and evaporation. Water quality remained safe and stable. No fish died. Overall
average weight gain was 43.9% for fish and 169.0% for plant. 24-h light resulted in 2.4% higher fish
growth, 12% higher plant growth and lower accumulation of all nitrogen and phosphate species in water
than 12-h light. Increased feeding frequency favored stable and good water quality and fastened fish
growth and plant growth by as much as 4.9% and 11%, respectively.
Ó 2013 Elsevier Ltd. All rights reserved.

Keywords:
Fish waste water
Aquaponics
Photoperiod
Feeding frequency
Tilapia
Water spinach

1. Introduction
Aquaculture is the culture of aquatic organisms, commonly
referred as animals, in a designated water body. The water needs to
be treated whenever toxicants in it have built up beyond animal’s
safe level. Toxicants such as ammonia and nitrite are derived from
decomposition of unconsumed feed and metabolites or waste of
the animals. Hydroponics is the culture of aquatic plants in soilless
water where nutrients for plant’s growth come entirely from a
formulated fertilizer. Aquaponics (a portmanteau of the terms
aquaculture and hydroponics) integrates aquaculture and hydroponics into a common closed-loop eco-culture where a symbiotic
relationship is created in which water and nutrients are recirculated and reused, concomitantly fully utilized and conserved. In
aquaponics system, waste organic matters from aquaculture system, which can become toxic to animals, are converted by microbes
into soluble nutrients for the plants and simultaneously, hydroponics system has already treated the water and recirculates back
to aquaculture system with cleansed and safe water for the animals.
Besides its ecological merits, aquaponics system can obtain extra


* Corresponding author. Department of Aquaculture, National Taiwan Ocean
University, Keelung, Taiwan. Tel.: þ886 2 24622192x5204; fax: þ886 2 24625393.
E-mail address: (Y.-H. Chien).
0964-8305/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
/>
economic advantages: saving cost (input) on water treatment for
aquaculture system, saving another cost on formulated fertilizer for
hydroponics system and benefit from double outputs, harvest of
animal and plant, by a single input, fish feed.
Tilapia is the most commonly used fish in aquaponics systems
(Rakocy et al., 2006) for their high availability, fast growing, stress
and diseases resistant and easy adaptation to indoor environment
(Hussain, 2004). Water spinach or swamp cabbage Ipomoea aquatica
is a semiaquatic tropical macrophyte and commonly grown as a leaf
vegetable in East and Southeast Asia. It has hollow stems, rooting at
the nodes and flourishes naturally in waterways or moist soil. It
requires little care to grow and hence low cost and popular in
Taiwan. It has been found effective in treating aquaculture waste
water (Li and Li, 2009) and eutrophic water with undesirable levels
of nitrogen and/or phosphorus (Hu et al., 2008). Nutrients dynamics
are quite complex in aquaponics system (Seawright et al., 1998). In
such system, feed is the primary source of nutrients which are
eventually tied up as the biomass of animal, plant and microbes or
stayed free in water. When no discharge, no nutrients are output
until the animal and plant are harvested as economic crops. Through
microbial decomposition, the insoluble fish metabolite and unconsumed feed are converted into soluble nutrients which then can
be absorbed by plant. Therefore, plant growth and production are
indirectly related to feeding strategies, fish metabolic condition and
microbial activity. While plant removes the soluble nutrients, water



694

J.-Y. Liang, Y.-H. Chien / International Biodeterioration & Biodegradation 85 (2013) 693e700

is filtered. Consequently, water quality or safe guard of fish growth
and production depends highly on the disposal capacity of the plant.
Besides the above factors which affect the nutrient availability
for plant and fish, system designs, plant and fish species and other
physical factors such as temperature, light sources and photoperiod
all add up the managing complexity for a steady state of nutrient
flow, which can be essential for the stable and predictable production of fish and plant in aquaponics system. Since photoperiod
affects photosynthesis and plant growth, the increase of photoperiod may also increase the removal capacity of nutrients in aquatic
macrophyte. Some studies have showed that the growth and productivity of floating aquatic macrophytes are directly related to the
intensity and amount of light, so are the absorption rates for nutrients (Gopal, 1987; Urbanc-Bercic and Gaberscik, 1989). Petrucio
and Esteves (2000) found that 2 h longer photoperiod a day
enabled two aquatic macrophytes to reduce more nutrients from
the water. Feeding frequency can affect feed intake of fish, quantity
of uneaten feed, feed utilization efficiency and consequently,
metabolite and excreta of fish and water quality. In an intensive
culture of fingerling walleye Stizostedion vitreum, Phillips et al.
(1998) found that higher frequency feeding resulted in higher
daily dissolved oxygen (DO) and lower total ammonia nitrogen.
Postlarval Ayu Plecoglossus altivelis with higher feeding frequency
at lower feeding rate had higher survival and growth (Cho et al.,
2003). When fed at 10% body weight daily, newly weaned Australian snapper Pagrus auratus fed 8 times a day had higher growth
and lower size heterogeneity than fed 4 and 2 times a day (Tucker
et al., 2006). Therefore, in the present study we investigate under a
constant nutrient input, namely, the feeding rate, if increasing

photoperiod can increase plant production, concomitantly, plant’s
filtering capacity and nutrient concentration in water and also if
increasing feeding frequency can even out through time, fish
metabolite and excrete, concomitantly, stabilize water quality and
increase fish production.
2. Materials and methods
The experiment had a factorial arrangement of 6 treatments,
namely, 3 feeding frequencies for red tilapia Oreochromis sp. X 2
photoperiods for water spinach I. aquatica Forsk. Illumination was
12 h or 24 h daily. An equal daily ration was evenly fed to the fish 2,
4 or 6 times at 12-h, 6-h and 4-h interval, respectively. Each
treatment had 2 replicates or experimental units. The experiment
was completed in 4 weeks.
Each experiment unit had an orange plastic tank (115 cm L Â
102 cm W Â 99 cm H), filled with 1000 L freshwater and stocked
with 8 fish at 467 Æ 30 g each or around 3.7 kg mÀ3. Constant
aeration was provided at tank bottom by a membrane disc diffuser
(LTD-325/325 mm, Aquatek, Kaohsiung, Taiwan), which had a
membrane diameter of 32.5 cm and provided an intensive air
throughput of 0.02e0.12 CMM 1e3 mm diameter air bobbles. A
piece of 3-cm thick polyethylene raft covered almost entire water
surface except a 15 cm  15 cm corner cut open, allowing an
automatic feeder release pellet feed into the water. A cut plant stem
25.1 Æ 3.7 cm or 7.8 Æ 0.5 g was wrapped around with layers of
sponge, stuffed in a black plastic ring (4.5-cm D) then fit into one of
the 63 evenly distributed round holes. Total plant biomass on a raft
was 490.2 Æ 5.5 g. Part of a stem was submerged to expose its first
bottom node, allowing for root initiation. A piece of coarse screen
(2.54 cm mesh) was secured 20 cm below the polyethylene raft to
prevent the plant root from possible disturbance by the fish. Each

tank was encased in a 200-cm tall wooden framework, which a
timer, a feeder and an illumination device could be fixed onto. A
near-sunlight 28-W 115-cm T5 tube was used for illumination,
hanging 25 cm above plant top and its height was adjusted as the

plant grew. Top and sides of the framework were covered with
black vinyl to obstruct the interference of illumination from
outside.
Each day same ration of feed for all experimental units was hand
loaded in the funnel of a feeder. Coupled with a timer, the feeder
released feed 2, 4 or 6 times a day at 12 h, 6 h or 4 h interval,
respectively, as designated in the experimental design. In this 4
week period, daily ration was gradually decreased from 5% to 3%
fish biomass as fish grew. A commercial tilapia feed was used,
which contained 25% crude protein, 3% crude fat, 12% crude ash,
6.5% crude fiber, 2% acid insoluble and 11% moisture. No water was
added or exchanged throughout the experiment.
Water was sampled weekly and monitored for pH (HM-20P,
DKK-TOA, Tokyo), dissolved DO and temperature (Oxi 330i, WTW
GmbH, Weilheim, Germany) and electrical current (EC) (750II
conductivity/TDS monitor, Myron L. Company, Carlsbad, CA). Total
ammonia-N (TAN), nitrite-N, nitrate-N, total nitrogen, soluble
phosphate-P, total phosphorus were analyzed by flow injection
analyzer (FIA) (Flow SolutionÔ FS3100, O. I. Analytical, College
Station, TX). The absorbance wavelengths used and methods based
for the analysis of those substances were 640 nm and phenol hypochlorite method (Solorzano, 1969) for ammonia nitrogen,
543 nm and Pink azo dye method (APHA, 1992) for nitrite nitrogen,
543 nm and CdeCu reduction method (Bendschneider and
Robinson, 1952; Strickland and Parsons, 1972; APHA, 1992) for nitrate nitrogen, 543 nm and CdeCu reduction method (Strickland
and Parsons, 1972; APHA, 1992) by Grasshoff et al. (1983) for total

nitrogen, 885 nm and molybdenum blue method (Strickland and
Parsons, 1972) for phosphate and 885 nm and method by
Grasshoff et al. (1983) and Strickland and Parsons (1972) for total
phosphorus. Analysis of five days’ Biochemical Oxygen Demand
(BOD5) was based by Sawyer et al. (2003). Biomass of fish and plant
was measured at 0, 2 and 4 wk. In wk 2, plant 25 cm above the raft
was cut, weighed and harvested. In wk 4, all fish and whole plant
were harvested. Fish weight gain (%) was calculated as the ratio of
average individual fish weight in wk 2 or wk 4 to average individual
fish weight in wk 0. Plant weight gain (%) in wk 2 was calculated as
the ratio of the biomass of cut part/initial biomass and plant weight
gain (%) in wk 4 further added the ratio of the biomass of whole
plant/initial biomass.
Three-way ANOVAs were performed to determine time effect,
the effects of photoperiod and feeding frequency and their interaction on fish and plant growth and water parameters. Duncan’s
multiple range test (DMRT) was used to compare differences
among levels of a factor. The significant level applied to all analyses
was set to 5%. SAS version 9.0 software (SAS Institute, Inc., Cary, NC)
was used for statistical analysis.
3. Results and discussion
3.1. System setup (Fig. 1)
Raft aquaponics can be the most simple and least cost aquaponics system. The essential elements of an aquaponic system as
suggested by Rakocy et al. (2006) are fish-rearing tanks, a settleable
and suspended solids removal component, a biofilter, a hydroponic
component and a sump. In raft aquaponics if the plant and its
supporting media such as gravel and coarse sand can provide sufficient biofiltration, a separate biofilter is not needed (Rakocy et al.,
2006). Solids removal component is highly recommended by
Rakocy et al. (2006) since if otherwise the organic materials from
fish fecal waste and unconsumed feed accumulate, deposit and
decompose anaerobically in tank bottom, the reduced toxic products can deteriorate water and harm the fish. In our system, the

upwelling afloat from disk membrane diffuser kept the solids


J.-Y. Liang, Y.-H. Chien / International Biodeterioration & Biodegradation 85 (2013) 693e700

695

Blidariu and Grozea (2011): an aquaculture system that incorporates the treatment and reuse of water with less than 10% of
total water volume replaced per day. Water replacement can vary
with the system setup. In a Nile tilapia and lettuce aquaponics
system where fish culture tanks, aquaponic channels, netting tanks,
clarifier and sump were open to evaporation, 1.4% of the total system water was added daily to compensate the evaporation and
transpiration losses (Al-Hafedh et al., 2008). The water consumption for fish production in the present study was 0.020 m3 kgÀ1 or
20.0 L kgÀ1 when calculating from the following data on per tank
base: initial fish biomass 3.7 kg, overall average fish weight gain
43.9%, initial water volume 1000 L, water loss 3.3%. The other water
consumption data from previous studies are as the following:
extensive fish culture >5 m3 kgÀ1 and semi-intensive fish culture
2.5 m3 kgÀ1 (both cited by Al-Hafedh et al., 2008), aquaponics by AlHafedh et al. (2008) 0.32 m3 kgÀ1 and aquaponics by Rakocy et al.
(1997) 0.25 m3 kgÀ1.

Fig. 1. A tilapiaewater spinach raft aquaponics system setup.

suspended and the effective aeration from the diffuser could
mineralize the organic particles so that little deposition was
observed in tank bottom at the end of experiment.
The level of polyethylene raft descended average 3.3 cm so that
the water loss in 4 wk was estimated only about 3.3% or about
0.1% dÀ1. Since water surface was afloat with polyethylene raft, only
the feeding corner area (15 cm  15 cm) was open to evaporation.

Other part of the water loss could be attributed to transpiration at
the leaf surface, which could be quite limited, too. Aquaponic system is a recirculating aquaculture system (RAS) as defined by

3.2. Treatment effects on fish survival and growth and plant growth
(Fig. 2)
Both feeding frequency (FF) (no. of meals dÀ1) and photoperiod
(PP) (illumination h dÀ1) had no effect on fish mortality since no
fish died throughout the experiment. Increased FF favored fish
growth since in wk 2 fish weight gain (WG) for 6 dÀ1 was already
1.2% and 2.2% higher than WG for 4 dÀ1 and 2 dÀ1, respectively, but
there was no difference between WG for 4 dÀ1 and for 2 dÀ1. FF
effect on growth became even more pronounce in wk 4 that 6 dÀ1
had 3.0% and 4.9% more WG than 4 dÀ1 and 2 dÀ1, respectively.
Generally, fish growth increases with feeding frequency. In indoor,
intensive fish culture systems, fish may be fed as many as 5 times
per day in order to maximize growth at optimum temperatures

Feeding frequency
2 d-1
4 d-1
6 d-1

Photoperiod
12h d-1
24h d-1

b
41.6
(1.9)


50

b
43.5
(1.9)

a
46.5
(0.8)

40

30

b
20.0
(0.9)

b
21.0
(0.9)

a
22.2
(0.4)

20

60


Fish weight gain (%)

Fish Weight gain(%)

60

a
45.1
(2.3)

b
42.7
(3.5)

50

40

30

a
21.6
(1.1)

b
20.5
(1.7)

20


10

10

2

4

180

c
164
(8)

160

140

c
133
(4)

b
134
(4)

a
138
(6)


b
169
(9)

a
175
(8)

Plant weight gain(%)

Plant weight gain (%)

200

b
163
(6)

b
131
(2)

a
175
(4)

a
139
(3)


120

Week

Week

Fig. 2. Mean value and standard deviation (in parenthesis) of fish weight gain (upper) and plant weight gain (bottom) of the tilapiaewater spinach aquaponic system under the
effects of feeding frequency (n ¼ 4) (left) and photoperiod (n ¼ 6) (right) during two sampling periods. Mean values without sharing a common letter are significant different
(p 0.05).


696

J.-Y. Liang, Y.-H. Chien / International Biodeterioration & Biodegradation 85 (2013) 693e700

Feeding frequency
4 d -1
6 d -1
2 d -1

Photoperiod
24h d -1
12h d -1

Total nitrogen (mgL-1)

a
a
b
46.0 b

44.7
38.6 b (4.6) 40.9 c
a
b ab (5.8)
(7.8)
(10.0) 37.3
38.8 35.7 36.6
37.4
(1.3) (2.8) (5.1)
(9.9)
(13.6)

a
a
25.4
23.4
(2.5)
(0.7)

a
a
a
25.4 23.8 24.1
(1.0) (0.9) (0.4)

b
35.1
(6.2)

b

34.1
(5.3)

a
a
6.5 6.5
(0.1) (0.1)

a
a
a
6.5 6.5 6.5
(0.1) (0.1) (0.1)

0

1

2

3

4
50

b
b
33.5 32.0
(6.1) (7.9)


a
a
a
21.5 21.3 20.6
(1.3) (0.8) (0.7)

a
a
a
0.1 0.1 0.1
(0.1) (0.1 (0.1)

a
20.5
(0.7)

30

a
39.9
(2.3)

a
40.0
(0.4)

a
a
33.6
31.7

(2.5)
(3.0)

40

Nitrate-N (mgL-1)

a
38.5
(5.2)

a
36.7 ab b
a
a
a
33.5
33.0 32.4 32.6 (3.6)
32.0
(6.2)
(3.2) (2.6) (3.6)
(7.9)

Nitrate-N (mgL-1)

a
47.8
(1.7)

a

46.3
(3.2)

a
38.9 b
(1.9) 35.1
(3.6)

b
30.1
(6.7)

a
20.9
(0.7)

20

10

a
0.1
(0)

a
0.1
(0.1)

0
0


1

2

3

4

Week

Week

Fig. 3. Mean value and standard deviation (in parenthesis) of total nitrogen (upper) and nitrate-N (bottom) concentration in water of the tilapiaewater spinach aquaponic system
under the effects of feeding frequency (n ¼ 4) (left) and photoperiod (n ¼ 6) (right) during two sampling periods. Mean values without sharing a common letter are significant
different (p 0.05).

Photoperiod
24h d -1
12h d -1

8

8

7

7

5


a
3.2
(2.2)

4
3
2
1

a
a a
0.1 0.1 0.1
(0) (0) (0)

a
a
a
0.8 0.7 0.8
(0.5) (0.2) (0.3)

a
5.0
(1.8)

a
4.7
(2.2)

6


ab
3.1
(3.1)

b
2.0
(2.4)

a
1.1
(0.6) a
0.7
(0.4)

Ammonia-N (mgL-1)

Ammonia-N (mgL-1)

Feeding frequency
4 d -1
6 d -1
2 d -1

ab
3.0
(2.8) b
2.2
(1.4)


a
5.2
(1.7)

6

5
4
3

2
1

a
0.1
(0)

a
0.1
(0)

a
0.9
(0.2)

a
0.6
(0.1)

a

1.8
(1.6)

a
1.5
(1.9)

b
1.3
(1.2)

b
1.6
(1.8)

0

0
0

1

2

3

4
1.6

a

1.3
(0.4)

a
a
a
0.6
0.5 0.6
(0.2)
(0.2)
(0.2)

a
0.8
(0.2)

a
0.6 a
(0.2) 0.5
(0.1)

a
a
a
0.1 0.1 0.1
(0) (0 (0)

1.4

a

0.8
a
(0.7) 0.7
(0.2)

a
0.8
(0.6)
a
0.5
(0.7)

a
0.5
(0.2)

Nitrite-N (mgL-1)

Nitrite-N (mgL-1)

a
6.2
(0.7)

1.2
1.0
a
a
0.6 0.6
(0.1) (0.1)


0.8

a
a
0.6 0.6
(0.2) (0.2)

a
a 1.0
(0.2)
0.9
(0.1)

a
0.7
(0.2)

0.6
0.4
0.2

a
0.1
(0)

a
0.1
(0)


0.0
0

Week

1

2

3

4

Week

Fig. 4. Mean value and standard deviation (in parenthesis) of ammonia-N (upper) and nitrite-N (bottom) concentration in water of the tilapiaewater spinach aquaponic system
under the effects of feeding frequency (n ¼ 4) (left) and photoperiod (n ¼ 6) (right) during two sampling periods. Mean values without sharing a common letter are significant
different (p 0.05).


J.-Y. Liang, Y.-H. Chien / International Biodeterioration & Biodegradation 85 (2013) 693e700

(Craig and Helfrich, 2002). Riche et al. (2004) found that feeding
tilapia at intervals shorter than the time required for the return of
appetite can lead to gastric overload resulting in reduced absorption efficiency. The return of appetite following a satiation meal,
defined as the point that consumption is equivalent to the amount
of the previous meal evacuated, is approximately 4 h in Nile tilapia
held at 28  C. In the present study, the interval of the highest FF,
6 dÀ1 was no shorter than 4 h and the absorption efficiency should
not be reduced. In an intensive culture of fingerling walleye

S. vitreum, Phillips et al. (1998) found that higher frequency feeding
resulted in higher daily DO and lower TAN but had no effect on fish
growth and size distribution. In conclusion, higher FF with less feed
given at a time can result in higher absorption efficiency and lower
excretion into water, consequently, less nutrient accumulation in
water.
Long PP increased fish growth since full day illumination
(24 h dÀ1) resulted in 1.1% and 2.4% higher WG than half-day illumination (12 h dÀ1) in wk 2 and wk 4, respectively. The present
results was not consistent with the results of El-Sayed and
Kawanna (2004), who reared fingerling Nile tilapia Oreochromis
niloticus under four photoperiod (light:dark, L:D) cycles (24L:0D,
18L:6D, 12L:12D and 6L:18D) at same feeding rate and feeding
frequency for 90 days and found that fish performance was not
significantly affected by photoperiods. Since in all tanks the polyethylene raft blocked most illumination onto the water, the only
difference in illumination resulted from the two PP was that the
feeding corner of 24 h dÀ1 received twice illumination as 12 h dÀ1,
which might somewhat help fish’s feeding and then the growth. FF
and PP had no interaction effect on fish growth.
Plant partial harvest in wk 2 had already showed that fish FF
helped for plant growth since plant WG increased with FF, namely,

6 dÀ1 > 4 dÀ1 > 2 dÀ1 in both wk 2 and wk 4. In wk 4 while the
whole plant was harvest, 6 dÀ1 had 11% and 6% more WG than 2 dÀ1
and 4 dÀ1, respectively. Long illumination had positive effect on
plant growth since PP at 24 h dÀ1 obtained 8% and 12% higher WG
than 12 h dÀ1 in wk 2 and wk 4, respectively. It is comprehensible
that longer illumination resulted in longer photosynthesis and
faster plant growth but not comprehensible how higher FF can link
to better plant growth.
3.3. Treatment effects on nitrogen species (Figs. 3 and 4)

In total nitrogen (TN), nitrate-N contributed around 88%,
ammonia-N (TAN) 11% and nitrite-N < 1%. Regardless of the
treatments, the overall average TN, nitrate-N and TAN increased
markedly until wk 2, slightly wk 2 to 3 and leveled off wk 3 to 4,
showing their accumulation had lessened. The overall average
nitrite-N reached its peak in wk 3, 0.9 mg LÀ1 and decreased to
0.6 mg LÀ1 in wk 4. The ammonia (NH3) safe level for Nile tilapia is
0.42 mg LÀ1 (Stickney, 1979; Karasu Benli and Koksal, 2005). Since
this safe ammonia level is expressed in terms of free NH3, it has to
be transformed into TAN. When the NH3 fraction from TAN at
overall average pH 6.69 and temperature 29.6  C, 0.5% are
accounted for, the TAN safe level for Nile tilapia is 84 mg LÀ1, which
is far higher than the highest TAN in the present study. Therefore, it
can be concluded that our aquaponics system ammonia toxicity risk
free. In the present study, the highly oxidized environment made
nitrite-N, the transitional nitrogen species in nitrification process
unstable and in very low concentration, and probably insensitive to
the treatment effect throughout the experiment.
TN, nitrate-N and TAN decreased with increased FF since wk 2
or wk 3. The adverse effect of FF on nitrogen species was most

Photoperiod
12h d-1
24h d-1

a
a
70.7 71.6 a
a
(7.3) (4.8) 63.6

61.8 a
a
(11.0)
(10.6) 53.6 57.1
(3.6) (4.4)

a
a
a
15.6 12.6 14.8
(4.9) (3.5) (1.7)

a
36.4
b
(8.7) b
25.4
22.0
(5.6)
(1.2)

Total phosphorus (mgL-1)

Total phosphorus (mgL -1)

Feeding frequency
2 d-1
4 d-1
6 d-1


a
a
a
3.6 3.3 3.6
(0.1) (0.1) (0.1)

Week

Phosphorate (mgL-1)

Phosphorate (mgL-1)

a
a
a
3.4 3.3 3.4
(0.1) (0.1) (0.1)

a
a
a
22.4 18.2
20.4
a
a
(4.2)
a
(3.2) (1.2)
13.3
12.6

11.3
(3.8)
(1.1) (0.7)

a
ab
64.9 62.5 b
(6.6) (3.4) 56.7
(9.4)

a
73.2
(5.3)

a
59.7
(7.7)

a
a
3.5 3.6
(0.1) (0.1)
0

a
a
45.7 42.3 a
(7.6) (4.2) 40.7
(8.7)


697

a
a
15.5 13.1
(3.1) (2.7)

1

a
55.3
(6.5)

a
b
31.2
(9.8) 24.6
(5.8)

2

3

a
47.7
(5.5)

a
a
3.3 3.4

(0.1) (0.1)

a
a
12.8 12.0
(3.4) (2.6)

b
64.1
(8.4)

4

a
65.7
b
(4.8) 57.1
(6.8)
b
38.1
(4.0)

a
a
21.3 19.4
(3.9) (2.5)

Week

Fig. 5. Mean value and standard deviation (in parenthesis) of total phosphorus (upper) and phosphate (bottom) concentration in water of the tilapiaewater spinach aquaponic

system under the effects of feeding frequency (n ¼ 4) (left) and photoperiod (n ¼ 6) (right) during two sampling periods. Mean values without sharing a common letter are
significant different (p 0.05).


698

J.-Y. Liang, Y.-H. Chien / International Biodeterioration & Biodegradation 85 (2013) 693e700

pronounced on TN and TAN in wk 4 when their concentrations
were in the order of 2 dÀ1 > 4 dÀ1 > 6 dÀ1. FF showed no effect on
nitrite-N, possibly due to its low concentration, <1 mg LÀ1, and
relatively wide variation. Juvenile (3.0 Æ 0.2 g) gibel carp (Carassius
auratus gibelio) were fed to satiation for 8 weeks to investigate the
effect of feeding frequency (2, 3, 4, 12 and 24 meals per day) on
growth, feed utilization and size variation, the results showed that
apparent digestibility of protein and energy increased significantly
at high feeding frequencies (Zhou et al., 2003). Higher feeding
frequency leads to better apparent digestibility may explain lower
concentration of nitrogen excreted into water.
PP exhibited its effect on reducing TN, nitrate-N and TAN since
wk 2, wk 3 and wk 3, respectively. In wk 4, the reductions were
28.7% ((47.8 À 35.1)/47.8), 26.3% and 69.2% for TN, nitrate-N and
TAN, respectively. However, PP had no effect on nitrite-N. Various
nitrogen forms have different effects on growth and nitrogen uptake of plants. Several studies have shown that aquatic macrophytes grow well when NH4 þ is the main nitrogen source probably
because less energy is needed for NH4 þ uptake and assimilation
compared to NO3 À nutrition (Petrucio and Esteves, 2000;
Jampeetong and Brix, 2009). When four aquatic macrophytes were
supplied with different inorganic nitrogen treatments: NH4 þ and
NO3 À each alone and together, the NH4 þ uptake rate was still
significantly higher than the NO3 À uptake rate (Jampeetong et al.,

2012). Petrucio and Esteves (2000) found that a macrophyte Eichhornia crassipes under 12 h dÀ1 PP removed from water 7.4% more
TN and 4.1% more NH4 þ than under 10 h dÀ1 PP.
3.4. Treatment effects on phosphate species (Fig. 5)
Both total phosphorus (TP) and phosphate ðPO4 3À Þ increased
continuously. The overall average TP and PO4 3À were 27.9 mg LÀ1

and 20.4 mg LÀ1 in wk 2 and increased 146% and 201% to
68.7 mg LÀ1 and 61.4 mg LÀ1 in wk 4, respectively. FF had no effect
on TP throughout the experiment. FF was effective in reducing
PO4 3À until wk 4 when 6 dÀ1 had lower phosphate concentration
than 2 dÀ1. PP at 24 h dÀ1 resulted in lower TP than at 12 h dÀ1 in wk
2 and wk 4 and such lowering PO4 3À effect occurred in wk 3 and wk
4. Petrucio and Esteves (2000) found that a macrophyte E. crassipes
under 12 h dÀ1 PP removes 9.2% more TP and 9.9% more PO4 3À than
under 10 h dÀ1 PP.
3.5. Treatment effects on electric current and BOD5 (Fig. 6)
Dissolved nutrients are measured collectively as total dissolved
solids (TDS), expressed as ppm, or as the capacity of the nutrient
solution to conduct an electrical current (EC), expressed as
millimhos cmÀ1 (mmho cmÀ1). In the present study EC increased all
the time. Its overall average value was 376 mmho cmÀ1 in wk 0 and
increase 84% to 694 mmho cmÀ1 in wk 4. This increase trend could
be attributed to the increase of ions derived from mineralization of
accumulated organic matter. Increased FF reduced EC starting in wk
3. In both wk 3 and 4, ECs in 6 dÀ1 and 4 dÀ1 were lower than EC in
2 dÀ1 but were not different between them. PP also took effect in
wk 3 that 24 h dÀ1 had lower EC than 12 h dÀ1. In wk 4, 24 h dÀ1
reduced EC by 19%. Rakocy et al. (2006) considered that in an
aquaponic system, TDS remains 200e400 ppm or EC 0.3e
0.6 mmho cmÀ1 will produce good results in plant production. If

dissolved nutrients are steadily increasing and approach 2000 ppm
as TDS or 3.5 mmho cmÀ1 as EC, phytotoxicity can occur and
increasing the water exchange rate or reducing the fish stocking
rate and feed input will quickly reduce nutrient accumulation. In
the present study, the EC in wk 4 was around 700 mmho cmÀ1, of
which the level of dissolved nutrients was suitable for plant’s

a
a
a
429 403 410
(10) (29) (11)

a
49.9
(18.6)
a
a
a
26.3
(10.5) 18.4 20.1
(7.4) (10.5)
a
a
a
5.5 5.6 5.5
(0.1) (0.1) (0.1)

a
649 b

(43) 582 b
(53) 554
(100)

a
69.8
(24.6)
a
33.8
a
28.9 (15.6)
(13.3)

a
78.8
(23.8)

a
746
(70)

b
b
663 655
(99) (108)

a
ab
80.3 78.2
(15.0) (24.8)

a
b
62.2
60.5
(20.0)
(28.5)

Electrical current (µmho/cm)

a
a
a
376 376 376
(1) (2) (1)

a
a
510 467 a
(29) (39) 475
(17)

Photoperiod
24h d -1
12h d -1

a
a
375 376
(1) (1)


a
43.2
(21.3)

a
a
5.5 5.5
(0.1) (0.1)

Week

a
a
422 406
(25) (20)

a
a
500
468
(27)
(33)

a
649
(34)

a
762
(34)

b
541
(65)

a
a
24.5
(11.2) 18.7
(6.3)

b
614
(75)

a
88.3
(17.9)

a
80.6
(19.3)

BOD5 (mgL-1)

BOD5 (mgL-1)

Electrical current (µmho/cm)

Feeding frequency
4 d -1

6 d -1
2 d -1

a
59.8
(20.5)

b
57.7
(17.2)

a
31.9
(11.1)

Week

Fig. 6. Mean value and standard deviation (in parenthesis) of electrical conductivity (upper) and BOD concentration (bottom) in water of the tilapiaewater spinach aquaponic
system under the effects of feeding frequency (n ¼ 4) (left) and photoperiod (n ¼ 6) (right) during two sampling periods. Mean values without sharing a common letter are
significant different (p 0.05).


J.-Y. Liang, Y.-H. Chien / International Biodeterioration & Biodegradation 85 (2013) 693e700

growth and only 1/5 of the concerned EC, 3.5 mmho cmÀ1, showing
that our aquaponics operated considerably satisfactory.
The increasing trend of BOD5 lessened in wk 3 when overall
average BOD5 in wk 3, 70.2 mg LÀ1 was not different from that in wk
4, 73.0 mg LÀ1. Only until wk 4 when FF had effect on BOD5 that
6 dÀ1 had 24% lower BOD5 than 2 dÀ1. Also only until wk 4 PP had

effect on BOD5 that 24 h dÀ1 resulted in 35% less BOD5 than 12 h dÀ1.
Among several conditions that biofilter performs optimally the
BOD5 should be <20 mg LÀ1 (Rakocy et al., 2006). In the present
study the raft aquaponics system did not equip with a separate or
independent biofilter, but had water spinach’s root cluster and
polyethylene raft’s undersurface functioned as a biofilter, of which
the BOD5 might not be subjected to the limitation of <20 mg LÀ1
since BOD5 in all treatments were already close to 20 mg LÀ1 in wk 2
and much greater afterward. Less fish excretion in higher FF can
be the reason for lower BOD5. PP at 24 h dÀ1 might lead to
higher filtering function in plant root and resulted in lower BOD5
than PP at 12 h dÀ1. However, why until wk 4 this happened was not
known.
3.6. Treatment effects on dissolved oxygen, pH and temperature
The overall average water temperature was 29.6 Æ 0.1  C. There
was no trend on temperature change with a maximum 30.1  C in
wk 2 and minimum 29.4  C in wk 4. FF did not affect temperature as
expected. Full day illumination neither result in higher temperature
than half day illumination the possible reasons could be: (1) the T5
tube’s fluoresce light does not generate much heat and the tube
itself seldom reaches 37  C as the specification claims, therefore,
not much difference in heat gain from longer illumination and (2)
polyethylene raft blocked the water from receiving light and itself
was a thermo insulator.
The overall average pH was 6.7 Æ 0.3. Weekly average pHs were
7.2, 7.3, 6.4, 6.4 and 6.2 for wk 0 to 4, respectively. A significant pH
drop during wk 1 to wk 2 could be coincided with the measurement right after the plant’s partial harvest. Speculation for such pH
change is not intended here since similar gap has not been
observed or evident in the other water parameters. No treatment
effects, neither FF nor PP, were found.

The overall average dissolved oxygen (DO) was 6.64 Æ
0.25 mg LÀ1, which was 86e88% saturated since 100% saturation
level at water temperature 29e30  C was 7.67e7.54 mg LÀ1. It is
suggested that DO should be maintained >5 mg LÀ1 for aquaculture
(Boyd, 1992; Graber and Junge, 2009). Aeration provided all DO in
this system and the membrane disc diffuser had demonstrated its
efficiency here. No time and treatment effects on DO were found.
Rakocy et al. (2006) addressed the importance to maintain high DO
levels in aquaponics systems for the health of root (Goto et al., 1996)
and the elimination of reduced toxicants in water.
4. Conclusion and recommendation
The present study demonstrated that the tilapiaewater spinach
raft aquaponics we used in this study was extremely effective in
fish waste treatment and also water conserved. In 4 weeks’ production period, only 3.3% water was lost due to vegetable leaf
transpiration and minor evaporation. Consequently, water consumption for fish production was exceptionally low, 20.0 L kgÀ1. No
fish died. Overall average weight gain was 43.9% for fish and 169.0%
for plant. Water quality remained safe and stable. Furthermore, as
expected the extending photoperiod and increasing feeding frequency increased both fish and plant production and lessened the
accumulation of nitrogen and phosphorus nutrients in water. These
findings are valuable and applicable in the development of aquaponics and biological waste reuse.

699

Current work can be fine-tuned to increase feeding frequency to
8 dÀ1 and 12 dÀ1 and add two more illumination levels between
24 h dÀ1 and 12 h dÀ1, namely, 20 h dÀ1 and 16 h dÀ1 so that a model
for optimal feeding frequency and photoperiod can be constructed
to further improve the efficiency. Researches are worth to explore
on the effects of various Illumination regimes, such as light intensity and light sources, i.e., different color LED (light-emitting
diodes), T5, incandescent lamp, halogen lamp, fluorescence lamp

and so on for most efficient illumination in energy saving, plant
growth, water treatment and fish growth.

References
APHA (American Public Health Association), 1992. Standard Methods for Examination of Water and Wastewater, eighteenth ed. American Public Health Association, American Water Works Association and Water Pollution Control
Federation, Washington, D.C., USA.
Al-Hafedh, Y.S., Alam, A., Beltagi, M.S., 2008. Food production and water conservation in a recirculating aquaponic system in Saudi Arabia at different ratios of
fish feed to plants. Journal of World Aquaculture Society 39, 510e520.
Blidariu, F., Grozea, A., 2011. Increasing the economical efficiency and sustainability
of indoor fish farming by means of aquaponics e review. Animal Science and
Biotechnology 44, 1e8.
Bendschneider, K., Robinson, R.J., 1952. A new spectrophotometric determination of
nitrite in seawater. Journal of Marine Research 2, 87e96.
Boyd,, C.E., 1992. Water Quality in Ponds for Aquaculture. Alabama Agricultural
Experimental Station, Auburn University, Auburn, Ms.
Cho, S.H., Lim, Y.S., Lee, J.H., Lee, J.K., Park, S., Lee, S.M., 2003. Effects of feeding
rate and feeding frequency on survival, growth, and body composition of Ayu
post-larvae Plecoglossus altivelis. Journal of World Aquaculture Society 34,
85e91.
Craig, S., Helfrich, L.A., 2002. Understanding Fish Nutrition, Feeds, and Feeding.
Virginia Cooperative Extension Publication 420-256. Virginia Polytechnic
Institute and State University, Petersbury, VA, p. 5.
El-Sayed, A.M.E., Kawanna, M., 2004. Effects of photoperiod on the performance of
farmed Nile tilapia Oreochromis niloticus: I. Growth, feed utilization efficiency
and survival of fry and fingerlings. Aquaculture 231, 393e402.
Gopal, B., 1987. Water hyacinth. Elsevier Aquatic Plant Studies 1, 471.
Goto, E., Both, A.J., Albright, L.D., Langhans, R.W., Leed, A.R., 1996. Effect of dissolved
oxygen concentration on lettuce growth in floating hydroponicsProceedings of
the International Symposium in Plant Production in Closed Systems. Acta
Horticulture 440, 205e210.

Graber, A., Junge, R., 2009. Aquaponic systems: nutrient recycling from fish
wastewater by vegetable production. Desalination 246, 147e156.
Grasshoff, K., Ehrhardt, M., Ehrhardt, K., 1983. Methods of Seawater Analysis. Verlag
Chemie, Federal Republic of Germany, p. 419.
Hussain, M.G., 2004. Farming of Tilapia e Breeding Plans, Mass Seed Production and
Aquaculture Techniques. Bangladesh Fisheries Research Institute, Mymensingh
2201, Bangladesh, p. 149.
Hu, M.H., Ao, Y.S., Yang, X.E., Li, T.Q., 2008. Treating eutrophic water for nutrient
reduction using an aquatic macrophyte (Ipomoea aquatica Forsskal) in a deep
flow technique system. Agriculture Water Management 95, 607e615.
Jampeetong, A., Brix, H., 2009. Nitrogen nutrition of Salvinia natans: effect of
inorganic nitrogen form on growth, morphology, nitrate reductase activity and
uptake kinetics of ammonium and nitrate. Aquatic Botany 90, 67e73.
Jampeetong, A., Brix, H., Kantawanichkul, S., 2012. Effectsof inorganic nitrogen
forms on growth, morphology, nitrogen uptake capacity and nutrient allocation
of four tropical aquatic macrophytes (Salvinia cucullata, Ipomoea aquatica,
Cyperus involucratus and Vetiveria zizanioides). Aquatic Botany 97, 10e16.
Karasu Benli, A.C., Koksal, G., 2005. The acute toxicity of ammonia on tilapia
(Oreochromis niloticus L.) larvae and fingerlings. Turkish Journal of Veterinary
Animal Science 29, 339e344.
Li, W., Li, Z., 2009. In situ nutrient removal from aquaculture wastewater by aquatic
vegetable Ipomoea aquatic vegetable Ipomoea aquatica on floating beds. Water
Science and Technology 59, 1937e1943.
Petrucio, M.M., Esteves, F.A., 2000. Influence of photoperiod on the uptake of nitrogen and phosphorus in the water by Eichhornia crassipes and Salvinia
auriculata. Review of Brasilia Biology 60, 373e379.
Phillips, T.A., Summerfelt, R.C., Clayton, R.D., 1998. Feeding frequency effects on
water quality and growth of walleye fingerlings in intensive culture. Progressive
Fish-Culturist 60, 1e8.
Rakocy, J.E., Bailey, D.S., Shultz, K.A., Cole, W.M., 1997. Evaluation of a
commercial-scale aquaponic unit for the production of tilapia and lettuce.

In: Fitzsimmons, K. (Ed.), Tilapia Aquaculture: Proceedings of the Fourth
International Symposium on Tilapia in Aquaculture at Orlando, Florida.
Northeast Regional Agricultural Engineering Service, Ithaca, New York, USA,
pp. 357e372.
Rakocy, J.E., Masser, M.P., Losordo, T.M., 2006. Recirculating Aquaculture TankProduction Systems: Aquaponics e Integrating Fish and Plant Culture. Southern
Regional Aquaculture Center Publication no. 454, Stoneville, Ms, p. 16.


700

J.-Y. Liang, Y.-H. Chien / International Biodeterioration & Biodegradation 85 (2013) 693e700

Riche, M., Haley, D.K., Oeetker, M., Garbrecht, S., Garling, D.L., 2004. Effect of feeding
frequency on gastric evacuation and the return of appetite in tilapia Oreochromis niloticus (L.). Aquaculture 234, 657e673.
Sawyer, C.N., McCarty, P.L., Parkin, G.F., 2003. Chemistry for Environmental Engineering and Science, fifth ed. McGraw-Hill, New York.
Seawright, D.E., Stickney, R.R., Walker, R.B., 1998. Nutrient dynamics in integrated
aquaculture-hydroponics systems. Aquaculture 160, 215e237.
Solorzano, L., 1969. Determination of ammonia in natural waters by the phenol
hypochlorite method. Limnology and Oceanography 14, 700e801.
Stickney, R.R., 1979. Principles of Warmwater Aquaculture. John Wiley and Sons Inc.,
New York, NY, USA, p. 375.

Strickland, J.D.H., Parsons, T.R., 1972. A practical handbook of seawater analysis.
Fishery Research Board of Canada Bulletin 167, 310.
Tucker, B.J., Booth, M.A., Allan, G.L., Booth, D., Fielder, D.S., 2006. Effects of photoperiod and feeding frequency on performance of newly weaned Australian
snapper Pagrus auratus. Aquaculture 258, 514e520.
Urbanc-Bercic, O., Gaberscik, A., 1989. The influence of temperature and light intensity on activity of water hyacinth (Eichhornia crassipes (Mart.) Solms.).
Aquatic Botany 35, 403e408.
Zhou, Z., Cui, Y., Xie, S., Zhu, X., Lei, W., Xue, M., Yang, Y., 2003. Effect of feeding
frequency on growth, feed utilization, and size variation of juvenile gibel carp

(Carassius auratus gibelio). Journal of Applied Ichthyology 19, 244e249.