Technical Report (CARD 023/06 VIE)

Improvement of water quality of outflows from ponds to waterways Cao van Phung
1
and Bell R.W.
2

1. Cuu Long Rice Research Institute, O’Mon, Cantho Province, Vietnam. Email:
[email protected]
2. School of Environmental Science, Murdoch University, Murdoch 6150, Australia. Summary
In the Mekong Delta of Vietnam, water quality is critically important for human
health and wellbeing as well as ecosystem function. Catfish farming in the Mekong
Delta, which is mostly confined to the provinces of An Giang, Cantho, Dop Lap and
Vinh Long, is considered to be a significant source of water pollution in those
provinces. A survey of 240 catfish farmers in An Giang and Cantho provinces, and of
Clarias fish farmers in Cantho Province confirmed that discharge of waste water and
solid waste from these operations is directly to public water sources in canals and
rivers in over 75 % of cases. Water consumers in the areas where catfish ponds are
prevalent had a strong perception that catfish pond discharge was the major cause of
degraded water quality. Water sampling at 2 points of discharge in An Giang over a
5-month period confirmed that discharge causes canal water to exceed discharge
standards for COD, TSS and NH
4
. However, even background levels were generally
above the levels set for household use in Vietnam. Two case studies showed that

pollution caused by discharge from catfish ponds, from the perspective of farmers,
and also by field sampling at two locations.
Water quality sampling was then used to determine the effectiveness of wastewater
treatment by recycling it through rice padi fields. Finally, there is a discussion of the
principles that are expected to govern future use of waste water treatment through
discharge to rice padi fields.
Baseline study in Cantho and An Giang
A survey was conducted in October-November 2007 in Cantho city and January-
February, 2008 in An Giang province. In total 240 questionnaires completed by
stakeholders (rice and fish farmers were sampled in equal numbers) were collected (2
districts/province) (Table 1, 2). The baseline survey was reported fully in April 2009
(Cao et al. 2009). Key findings of the study are discussed below to highlight the main
water quality concerns, as perceived by the farmers. In addition, the survey reported
on the prevalence of fish and human diseases, which may be indicators of water
quality.
Water quality
Generally half or less of the farmers were satisfied with water quality, and 15-24 %
classified it as bad. Yet, only 5-8 % of fish farmers had their own settling pond and
the vast majority discharged waste to canals and rivers directly without treatment.
Phan et al. (2009) in a survey of 89 catfish farms along the main rivers of the Mekong
Delta found only 3 % of farmers reported having settling ponds. However, 15-24 % of
catfish farmers in the present study used rice padi fields for discharge of waste water.
Waste from fishpond was perceived to be the major source causing pollution in An
Giang while in Can Tho 37 % of respondents considered catfish ponds the major
cause for poor water quality. Part of the reason for the different responses may be that
50 percent of respondents in An Giang used river water for household purposes while
a similar proportion of people in Cantho utilized tubewell water (Cao et al. 2009).
Farmers were also concerned of threats to fish culture from pesticides discharges
from paddy fields.
The type of feed and additives supplied to catfish may have a bearing on water


46
39
15
Wastewater discharge
- River or canal
- Paddy
- Settling pond

80
15
5

100

68
24
8

100

Table 2: Reasons given for poor quality of water for irrigation and for household use
by catfish and rice farmers. Values are a percentage of respondents where N= 120 in
each province.
Items An Giang (%) Can Tho (%)
Waste from fishpond 91 37
Pesticides from paddy 9 15
Others 0 48

For water treatment in catfish culture, nearly 50 % of farmers at both sites had used

Mixing to feed 60 96
Mixing with fishpond water 40 4
Water treatment 97 100
BKC 21 28
Vikong 43 5
Copper sulfate 7 8
Chlorine 21 10
Others 8 49

All farmers normally spread lime or salt along the sides and bottom of ponds after
making new ponds or after draining out the water from harvested ponds. Dosages of
lime and salt varied from 300-400 kg/ ha. Pond are kept dried for 3-5 days before re-
filling with water for a new crop of fish. More than 95 % of farmers changed and/or
added water to fish ponds regularly (about 1/3 volume of pond daily) and there was
about 50 % farmers at both sites practiced bottom of fishpond cleaning by pumping
out sludge during the growth of catfish. In addition, all of the farmers pumped sludge
out of their pond after harvesting fish (Table 5). The practice of discharge of solid
waste directly to water sources was practiced by 63-70 % of catfish farmers (Table 6).
Most farmers had a handheld pH meter to monitor water quality (Table 5). Other
ways to detect water quality were by observation of water colour or smelling the
odour of water. Farmers conclude that oxygen is deficient if most of fishes come to
the pond surface during the early morning.
Table 5: Water management in catfish culture. Values are a percentage of respondents
where N= 60 in each province.
Items Cantho An Giang
Renewed fishpond water (%) 100 100
Lime (kg/ha) 425 350
Salt (kg/ha) 325 300
Dried bottom of pond (%) 100 100
Days of drying 3.5 4

water for catfish ponds from a common source. Hence disease organisms may spread
readily from one operator to another. The dispersion of disease-causing organisms in
the canals and river has not been investigated in the present study, although the
dispersion plume for TSS in waste water was at least 600 m from the point of
discharge. While the farmers reported that effective control measures for major
diseases were generally available, reducing the spread of the disease through better
water quality may be an outcome from more effective treatment of wastes from fish
ponds.

Table 6: Solid waste management on catfish farms. Values are a percentage of
respondents where N= 60 in each province.
Items Cantho An Giang
Removing sludge purpose

Sanitation 51 65
Disease control 46 35
Methods of discharge

Pumping sludge 86 89
Removing by hand 14 11
Location of sludge discharge

Water course 70 63
Filling up low lying surface 30
Mulching of fruit trees 37
Times of sludge discharge

One time 37 35
Two times 60 59
Three times 3 6

From 1-3 month old 17 8
More than 3 month old 11 4
Effective Control

Haemorrhage 92 82
Swelling head 77
Slimy loss 100

While not a rigorously collected set of data, the information on human disease trends
in two districts with a high concentration of catfish ponds may be related to the rapid
spread of fishponds in recent years.

Table 8: Disease reports to Preventative Medical Centre in 2006 & 2007 at O Mon
and Thot Not districts
Disease O Mon Thot Not
2006 2007 2006 2007
Dengue 179 237 246 546
Typhoid fever 2 0
Diarrhea 159 220 3210 2822
Dysentery 0 1 138 31

Table 9: Disease reports to Preventative Medical Centre in 2006 & 2007 at Phu Tan
and Chau Phu districts
Disease Phu Tan Chau Phu
2005 2006 2007 2005 2006 2007
Dengue 112 300 495 411 595 438
Typhoid fever 169 153 71
Diarrhea 752 957 1502 189 187 200
Dysentery 202 250 463 21 10



Figure 1: pH of water in canal before and after opening the flushing gate to discharge
fishpond waste water. Values are means of 40 values sampled at 15-day intervals for
5 months (December to April) in two sites of An Giang Province (Phu Tan and Chau
districts). Figure 2: Electrical conductivity (EC) of water in canal before and after opening the
flushing gate to discharge fishpond waste water. Values are means of 40 values
sampled at 15-day intervals for 5 months (December to April) in two sites of An
Giang Province (Phu Tan and Chau districts).

Normally COD of liquid waste from fish ponds exceeded values allowable for living
use (below 10 mg/L) except during the flooding time (October-January). Indeed, even
the values in the main canal before discharge of waste water were generally > 10 mg
/L. All values, before and after discharge were in the range accepted for water
discharge from fish ponds (35 mg/L <COD<100 mg/L).

0
50
100
150
200
250

Fisheries.
In the case of nitrite and nitrate levels taken at different time intervals before and after
opening the flushing gate, they were lower than allowable values for living use even
in the dry season when nitrite and nitrate levels in canal waters were more
concentrated than in the flooding season (data not shown).
On average, TSS values in canal waters after the opening the flushing gate (261± 64
mg/l) were five times higher than those recording before (54 ± 20 mg/l). Figure 5
demonstrated that most of the TSS values collected in canals under the effect of waste
water discharge were 10-13 times higher than TCVN 5942-1995 (column A- TSS less
than 20 mg/l). Even the allowable value (less than 80 mg/L) for waste from fish pond
discharged to water sources under Announcement No 02 of the Ministry of Fishery
was generally exceeded.
Hence these sampling results indicate clearly measurable effects of wastewater on
canal water quality for at least 100 m up-stream and 500 m downstream of the
discharge. The measurable decline in water quality was attributed to elevated COD,
NH
4
and TSS.
These water quality data support the results of the baseline survey of farmers. In
Cantho and An Giang, 7-36 % of farmers assessed water quality to be bad, and of
these 37 to 91 % attributed the poor water quality to catfish pond discharge. A higher
proportion of farmers were troubled by poor water quality in An Giang than in
0
20
40
60
80
0 100 200 300 400 500
Distance (m)
COD

Ammonia (mg/L)
After
Before
0
50
100
150
200
250
300
-100 0 100 200 300 400 500
Distance (m)
TSS (mg/L)
After
Before
in the tributary canals and in the closed-dyke system, measurable decline in water
quality is evident in the Phu Tan and Chau district sampling. Indeed, after discharge,
water sampled in canals was below the discharge standards for TSS, COD and NH4.
In the present study we did not study the longevity of these peaks in water quality to
ascertain whether they persist for minutes, hours or days. The persistence of the peaks
in water pollution due to fishpond discharge are likely to relate to the level of
pollution in the discharge water, and the volume of discharge in relation to water flow
in the canal. Small canals in the dry season with limited flow for flushing are most
likely to suffer declines in water quality.

Water quality in catfish ponds:
Investigations were conducted from December 2007 to July 2008 at Phu Tan and
Chau Phu districts of An Giang province on water quality in 12 catfish ponds that
used two types of feed, namely manufactured pellets and farm-made feed.
Results showed that pH values were still within the acceptance limit (7.0-8.0) even

(mg/L)
BOD
(mg/L)
COD
(mg/L)
TSS
(mg/L)
Pellet 0.145 0.032 0.134 0.662 0.416 13.3 69.4 86.9
Farm-
made 0.138
0.077 0.104 0.569 0.318
18.0
73.3
90.7
LSD
(5%) 0.017
0.012 0.012 0.246 0.093
1.94
14.5
15.3
Ponds using pellets had higher values of total available N and TP than farm-made
feed since the pellets were much more concentrated in nutrients (Table 10). However,
the nitrite and nitrate concentrations were lower in the ponds with pelleted feed
rations.
Total available N, nitrite, nitrate, phosphate, total N and total P increased steadily
with time after stocking the ponds and they were not much different between the two
sites (data not shown).

VAC water quality study
VAC systems are located mainly on high elevation areas along rivers. These are areas

2
of pond surface). Solid wastes are mixed with liquid and then discharged
to the environment. However, suspended particles will then settle along water ways
and the canals which obstructs navigation.
Table 11: General information on Clarias sp. Culture in two districts of Cantho
Province.
Items Binh Thuy Phong Dien
Type of fish culture (%)

Clarias sp. (%) 86 92
Others 14 8
Average time for fish raising (month) 3.5-5 3.5-5
From seed – fingerling (months) 1.2 1.3
From fingerling – selling size (months) 7.2 7.5
Size of fishpond (m
2
) 500-3000 1000-3000
Depth of fishpond (m)

Fingerling raising 1.2 1.1
Selling size 1.8 1.6
Density of fish/m
2
90-120 70-100
Conversion factor (feed/meat) 3.5 3.6

Table 12: Water quality of Clarias fishponds sampled in Binh Thuy and Phong Dien
districts of Cantho province.
Items Binh Thuy Phong Dien
BOD

Soil treatment (zeolite, lime) 62 75

Water quality after passage through padi fields
The effect of passage of fishpond waste water through padi fields on water quality
was assessed at two locations, Chau Phu and Phong Dien. In the former case the water
had first passed through a main canal receiving discharge from a large number of
catfish ponds. It was then directed through rice fields to another main canal. Water
samples were collected along this water course. In the second case, water was
sampled from a Clarias fishpond and then at a series of downstream locations to
determine changes in composition and quality.
In Chau Phu, water samples were collected along the course of water from a main
drain that received fishpond discharge, through small drains adjacent to rice padi
fields, settling ponds and small wetlands to discharge into another main canal (Fig. 6).
Water samples were analysed for ammonium-N, nitrate-N, total N, TSS, COD and
total P (Table 14).
During the passage of the water in Chau Phu from the fishpond-polluted main canal to
the discharge canal, levels of N and P dropped 3- to 4-fold (Table 14). The water
COD also dropped from over 230 mg /L to about 100 mg/L. There was however,
only a small decrease in TSS with the values at VTT10 and VTT11 still being about
100 mg/L and above both the standard for fishpond discharge (80 mg/L) and for
household use (20 mg/L).
In the settling ponds and wetland, the composition of water hardly changed compared
to that in the small drains upstream, indeed ammonium-N and nitrate-N increased. Fig. 6. Water sampling points in Chau Phu from the main canal (VTT1, VTT2- top
left), small field drains (VTT3,4,5), a settling pond (VTT6), a small wetland
(VTT7,8), a field drain (VTT9) and in the main canal (VTT10,11).

Water from Clarias fishponds at Phong Dien was directed through rice padi fields

VTT8 0.22 0.35 114 128 0.62 0.82
VTT9 0.06 0.08 112 82 0.34 0.28
VTT10 0.06 0.08 98 98 0.28 0.31
VTT11 0.09 0.05 104 64 0.32 0.34

Sites 13,14,15
Fig 7. Schematic representation of sampling locations: Clarias fishpond (Grey
rectangle- Sites 1,2,3); rice padi fields (green rectangle- Sites 4,5,6); canal (open oval
shapes- Sites 7,8,9 (upstream), Sites 10,11,12 (downstream)); discharge to main canal
(blue circle- Sites 13,14,15).
The Clarias fishpond water was highly turbid, excessively high in BOD and contained
high concentrations of total N, nitrate-N and total P (Table 15). The water after
passing through the padi field had decreased 15-fold in TSS and both the TSS and the
BOD had dropped to acceptable levels for discharge. Total N and nitrate-N had both
dropped 3 to 4-fold, while total P only dropped by 25 %. The composition of water
draining from the padi field was remarkably similar to that in the nearest canal, in
both upstream and downstream directions, as well as at the discharge to the nearest
main canal.

Table 15. Water quality in Clarias fishponds in Phong Dien (Sites 1-3), after passing
through padi fields (Sites 4-6) and in downstream canals (Sites 7-15). Values are
means of 3 sampling points at each location. See Fig. 7 for a schematic of sampling
locations.
Site pH
EC
(mS/cm)
TSS
(mg/L)
COD
(mg

need to be considered. The loading of waste to land depends on optimising four
criteria: hydraulic loading; solids loading; chemical reactivity of the waste, and;
nutrient assimilation capacity of the system. The optimisation of these factors in
relation to fishpond waste application to rice padi fields in the Mekong Delta is
discussed below.
Fishponds generate wastewater continuously throughout the year, while the solid
waste from the bottom of the fishponds is removed infrequently, usually at harvest of
the catfish after 6 months (Table 6). Application of both types of waste is considered
below. About half of the fishponds surveyed in Cantho and An Giang, periodically
flushed out solid waste during the raising of catfish (Cao et al. 2009; Phan et al. 2009)
which probably increases the nutrients, COD, BOD and TSS levels in the waste water
discharge. The concentration of nutrients and levels of COD and BOD are much
higher in solid waste than wastewater, but the large volume of wastewater means that
total nutrient loading may be as high. With very frequenct water replacement, the
composition of waste water may not be too different from that of canal water (Bosma
et al. 2009).
In the survey of Cantho and An Giang, fishpond operators replaced 33 % of the
fishpond water daily to avoid build up of high levels of nutrients and organic matter,
and to minimise levels of infection of fish by disease organisms (Cao et al. 2009). In
another study, the range was reported to be 20-40 % replacement daily (Phan et al.
2009), although the frequency is also reported to vary with the growth cycle of the
catfish. During the early growth of catfish, less frequent changes occur and at close to
maturity more frequent changes of water occur. The ponds with more frequent
changes probably also accumulate solid waste slowly. By contrast, when fishpond
operators change water less frequently, it will result in a wastewater with higher
nutrient and carbon loading, and a greater accumulation of solid waste at harvesting.
The stocking rate of fish and age of the fish (both of which affect the amount of feed
supplied per day) will also affect the composition of the wastewater. Use of bought
pelleted feed vs farmer-produced feed may also affect wastewater and solid waste
accumulation based on differences in feeding rates, the composition of the feed, and

timing of wastewater application to padi fields. Antecedent rainfall may block the use
of wastewater in a particular field for several days. Hence, it is important for the
flexibility of operations of fishponds to arrange surplus capacity of padi fields or
settling ponds on which to apply wastewater, to avoid the situation of having no
storage available for wastewater disposal when discharge is required from the
fishpond.
From a case study at Chau Phu, An Giang Province, it was reported that growing 3
rice crops per year required 12-13 irrigations, each supplying about 75 mm of water.
Hence in this location, 900-1000 mm of irrigation water was required per year. Many
locations only grow 2 rice crops per year, but in Chau Phu where 3 crops are grown
the main wet season crop uses no irrigation water. Hence, the above estimate in Table
18 of 835 mm annual irrigation requirement seems reasonable for either 3 crops or 2
crops, although perhaps a conservative estimate.
Large volumes of water need disposal from fishponds. Assuming an average depth of
3 m for fishponds (Phan et al. 2009) and a 9.2 cm depth of each waste water
application, 1 ha of fishpond requires about 33 ha of padi fields for each time the
pond is fully emptied and re-filled. This assumes rice uses 4.6 mm per day
(evaporation and transpiration- see Table 18) so that re-application of 9.2 mm
wastewater to a field needs to occur every 2 days. In practice farmers apply 50-100
mm of irrigation water every 10 days in the dry season and every 18-22 days in the
early wet season. The replacement of 33 % of pond water every day is a common
practice so full replacement notionally takes a 3-day cycle. Hence 10-20 ha of padi
land would be required to use the wastewater produced in the dry season from 1 ha of
fishponds assuming that only wastewater is used for irrigation. In the early wet
season, rainfall would prevent wastewater irrigation on some days, so that 20-40 ha of
padi land would be required for every 1 ha of fishpond.
Table 16. Example of monthly irrigation water required for rice fields in the Mekong
Delta and rainfall (averages for Long Xuyen and Cantho), and potential evaporation
(PE) from Ho Chi Minh City. Note that rice also receive water from groundwater
during daily high tides.

ha would be advisable to minimise the risks of nutrient build up as discussed above
and to allow for the unavailability of some fields at the time when the solid waste was
available.
2. Solid loading
The application of solids can cause "clogging" of soil pores at the soil surface, due to
the organic and inorganic particles applied. Clogging interferes with the movement of
gases and liquids into the soil matrix. In fishpond liquid waste, the level of suspended
solids is 200 mg/L (Fig. 5). Farmers at Chau Phu reported that irrigation with waste
water caused a slowing of infiltration rates. For rice, exact levelling was also required
to prevent waste water accumulating in low patches, where excessive N loading
caused a tendency to lodging of rice. For vegetables, slower infiltration rates were
observed when irrigating with waste water.
Solid waste contains much higher levels of solids. However, the padi rice system is
not a percolation based system of waste treatment, hence the blocking of surface
pores may not be a significant concern except before sowing of rice seeds on the
prepared seedbed. The need for precise field leveling would also apply with the use of
fishpond solid waste.
3. Chemical reactivity of the waste
Organic materials applied to land will produce a biochemical oxygen demand (BOD)
and a chemical oxygen demand (COD). Waste application rates should be designed so
that COD or BOD do not exceed the gas exchange capacity of the soil, otherwise
decomposition of the waste will be slowed by lack of oxygen, and anoxic conditions
will develop. Anoxia in the root zone is less of a concern for rice than for other crops.
A wide variation in BOD has been detected in different wastes (McGowan et al. 2007;
Table 17). The maximum recommended BOD for direct application is 100,000 mg/L
while maximum COD is 250,000 mg/L (Yadav et al. 2002). In fishpond waste water
in Vietnam, levels of BOD and COD were 35 and 100 mg/L, hence well below limits
for soil application.
Wastewater from a number of locations in An Giang and Cantho contained 0.5-5 mg
NH

fishponds are pumping mineral sediment from the bottom of ponds to increase their
depth. The composition at some sites in Table 20 may not reflect that obtained in the
long term when ponds achieve their planned depth. The composition of swine lagoon
solids reported by Duffera et al. (1999) suggests that the catfish pond solid waste is
relatively low in P and to a lesser extent in N.
Table 12: Nutrient composition in wastewater at experimental sites reported by Cao et
al. (2010) and Bosma et al. (2009). Except for the Phong Dien sites, all are water from
catfish ponds. Data reported by Bosma et al. (2009) was for 3 sites of undisclosed
location on the Mekong Delta.
Location pH EC
(µS/cm)
NH
4
-N
(mg/L)
NO
3
-N
(mg/L)
Total N
(mg/L)
Total P
(mg/L)
Total K
(mg/L)
CLRRI 7.18 373 0.46 2.72 5.39 2.53 4.18
Chau Phu 7.13 234 3.4 0.42 5.40 8.46 4.47
Phu Tan 7.32 243 4.84 0.79 7.66 6.44 5.12
Phong Dien 6.65 217 2.18 12.3 16.7 1.21 6.02
Mekong

Solid waste An
Giang 2007 mean of
12 farms
Solid waste An
Giang 2007
Swine lagoon
solids
C 5±0.3 % 2.5-2.7 % 13.7±0.4 %
N 0.32±0.03 % 0.83 % 1.72±0.04 %
C:N 15.6 3 8.0
P 0.28±0.08 0.61 % 2.49±0.11 %
K 1.1±0.03 3.5 % 0.67±0.03 %
Ca 0.013±0.006 % 7.8-8.8 %
Cu 35±1 32 mg/kg
Zn 111±4 109 mg/kg
Fe 3.9 ±0.3 % 3 %
Mn 0.05±0.01 % 331 mg/kg
pH 6.4-6.5
EC 1.9-2.1 mS/cm