Technical Report Irrigating rice crops with waste water to reduce environmental pollution from catfish production in the Mekong Delta Cao van Phung 1 , Nguyen be Phuc 2 , Tran kim Hoang 2 and Bell R.W. 3 1. Cuu Long Rice Research Institute, O’Mon, Cantho Province, Vietnam. Email: caovanphung@hcm.vnn.vn 2. An Giang University, Long Xuyen, An Giang Province, Vietnam 3. School of Environmental Science, Murdoch University, Murdoch 6150, Australia. Abstract Waste from intensive catfish (Pangasianodon hypophthalmus) aquaculture production has become a pollutant of surface waters in the Mekong Delta, Vietnam. In the present study, the aim was to treat the wastewater from catfish ponds in the Mekong Delta by land application to padi fields so that the nutrients could be recovered by rice crops as a fertilizer substitute. A survey in the dry season 2007 of paired fields in An Giang Province showed that rice yield in 16 paddies receiving waste from fishpond was 1 t/ha higher than in another 16 paddies that did not use wastes. In six field experiments using fishpond waste water for irrigation, decreasing fertiliser N by 33 % and P and K by 50 % had no effect on rice yields. In other cases, decreasing N by 40 % or P by 50 % did not decrease yield. The variation in nutrient composition in waste water among sites, and in yield potential and irrigation requirements especially  1 Corresponding author Cuu Long Rice Research Institute, O’Mon district, Cantho city-Vietnam. Phone No (84) 710861452. Fax: (84) 710861457. Email: caovanphung@hcm.vnn.vn between wet (lower yield potential and lower irrigation requirement) and dry seasons may account for the different extent of fertiliser replacement feasible by waste water without decreasing yield. Ten -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 may be required for every 1 ha of fishpond. If waste water application on this cycle causes excess nutrient loading on padi fields, a further increase in padi land is required to apply this approach as a sustainable strategy for treating fishpond waste water. Keywords: catfish, fishpond waste, nutrients, pollution, rice. Introduction Catfish culture in the Mekong Delta has been practiced for a long time but this industry became important for export only after the year 2000 with a subsequent annual growth rate up to 2008 of about 15-20 % per annum. Total catfish production in the Mekong delta was 0.68 million tonnes of fillets in 2007 (Phan et al. 2009). Cat fish production expanded to cover about 6,000 ha of ponds in the Mekong delta, Vietnam (Bosma et al. 2009). The production of each tonne of catfish consumes 4,023 m 3 of water and releases 47.3 kg of N (Phan et al. 2009). Wastewater discharged directly from intensive catfish aquaculture production is polluting surface waters in the Mekong Delta, Vietnam. From the catfish ponds, large quantities of liquid waste are discharged to waterways without treatment (Phan et al. 2009). It is estimated that about 2754GL of water was discharged annually back to the surface water system of rivers and canals of the Mekong Delta from catfish ponds. Consequently, the pollution of canals or rivers by loading of fishpond waste, rich in nutrients (especially nitrogen, phosphorus and carbon) has emerged as a major concern for sustainability of the industry (Phan et al. 2009). The introduction of the National Environment Law in 2005 which prohibits direct wastewater discharge of aquaculture water into rivers and canals, is an underlying driver for this study. Although enforcement and compliance with this regulation currently appears to be low, the future sustainability of fish pond aquaculture relies greatly on the ability of farmers to comply with environmental and export regulations. For this reason cost-effective wastewater treatment strategies need to be developed and applied by farmers. Presently only 15-24 % of catfish farmers in Cantho and An Giang provinces currently practice recycling of wastewater from their ponds by irrigating rice fields (Cao et al. 2009). Pollution due to fishpond waste is generally attributed to high organic carbon and nutrients (Pillay, 1992) although high total suspended solids, NH 4 -N and COD may also downgrade its acceptability for a range of uses. Moreover, this form of discharge is also contributing to the spread of diseases of catfish since downstream operators are pumping the infected water from canals into their ponds (Phan et al. 2009). Levels of fish-disease organisms in the river peak at the beginning of the rainy season in the Mekong Delta. The quantity of waste produced depends upon the quantity and quality of feed (Cowey and Cho, 1991). This is associated with lower feed conversion ratios from manufactured feed pellets than to farm made feed (1.69 vs 2.25; Phan et al. 2009). Hence the latter feed results in greater waste production. The frequency of water replacement and stocking densities in the fishponds will also affect the quantity and quality of waste water. However, integration of aquaculture into existing agricultural systems has been reported to improve productivity and ecological sustainability of both operations through better management and improved soil fertility arising from waste recycling (Bartone and Arlosoroff 1987). Moreover, properly managed inputs of waste materials can reduce the need for fertilisers (Falahi- Ardakani et al. 1987). Rice uses large volumes of water, especially in the dry season , when the crop is fully irrigated. Instead of using river or canal water to irrigate rice, wastewater, if available from adjacent catfish ponds, could supply most of the water requirements for rice while also providing a significant supply of nutrients. The present study aims at recycling waste water from fishponds by using it to irrigate rice. The rationale for the treatment was to use the assimilation capacity of rice to absorb nutrients and the filtering processes of the rice padi and associated water distribution canals and sediment traps to improve the quality of water discharged from fishponds before it re- enters the main canals and rivers. The objective of the present study was to determine the fertiliser substitution value of the fishpond waste water in order to determine how to adjust recommended fertiliser rates for rice when using this water to irrigate crops instead of river water. Materials and methods Field experiments on recycling of waste water were carried out on rice crops commencing with the wet season 2007 and ending with the dry season 2010. Site locations and soil characterisation are given in Table 1. A preliminary survey on the beneficial use of fishpond waste water for rice cultivation on farmers’ fields was carried out in the dry season 2007 at Chau Phu and Phu Tan districts of An Giang province. In each district, 16 fields were selected comprising 8 which used waste water from fishponds and the other 8 paired sites which were protected by levees to prevent inflow of waste water. Rice samples were harvested from 5 m 2 with 3 replications for yield evaluation. Experiments on recycling of waste water for rice production were carried out at two communes of Chau Phu district namely My Phu during wet season 2007 and dry season 2008 and at Vinh Thanh Trung for dry season 2009 and wet season 2009. Another 2 experiments were conducted during the dry season 2008 at Phu Tan district (two sites) of An Giang province. Further wastewater experiments were conducted at CLRRI in the dry season 2009 and at Phong Dieng in the dry season 2009. At the latter site, the wastewater was sourced from a Clarias fish pond, while for the remaining sites the water was from catfish ponds. Nutrient composition of wastewater at each site is shown in Table 2. There were 6 treatments for experiments at Chau Phu and Phu Tan using chemical fertilisers (N-P-K rates in kg/ha given in parentheses) as follows: T1 (90-26- 50); T2 (60-13-25); T3 (30-0-25); T4 (30-26-25); T5 (30-13-50) and T6 (0-13-50). Experiments in My Phu did not include T5. Inorganic fertilisers treatments for experiments at Vinh Thanh Trung and CLRRI were adjusted as follows: T1 (100- 17.5-25); T2 (80-14-25); T3 (60-10.5-25); T4 (40-14-25); T5 (40-10.5-25). Finally, experiments at Phong Dien tested two treatments viz, irrigation with wastewater and reduced inorganic fertiliser (45-13-30, NPK in kg/ha) vs irrigation with river water and the recommended fertiliser rate (83-21-17, NPK in kg/ha). Irrigation with wastewater occurrred 5 times for the wet season and 10 times for dry season rice crops. The volume of wastewater applied at each irrigation event was 1000 m 3 /ha (i.e. 10 cm depth of water). Nutrients composition of the wastewater at different sites is presented in Table 2. All experiment was laid out in randomized complete block design with 4 replications except Phong Dien which comprised 8 replications, each of 2 treatments. Plot size was 8 x 7 m and each plot was separated from others by bunds. Soil sampling on each plot was done before planting and after harvesting every crop. Yield components were estimated by sampling two 0.5 x 0.5 m quadrats on each plot. Actual yield was measured by harvesting 5 m 2 per plot. Rice crops were protected against leaf folder, thrips, brown plant hopper and rice blast by using effective pesticides as required. Organic carbon is determined by wet digestion; analysis of nutrients (N, P, K, Ca, Mg, Fe, Cu, Zn, Mn) followed standard methods for soil (Page et al. 1982), plant and water analysis (Chapman and Pratt, 1961). Statistical analysis was completed with IRRISTAT software version 5.1 by applying a balanced one-way ANOVA. Nutrient (N, P) balances were calculated following the approach of Dobermann and Fairhurst (2000) to estimate total input and output. Values reported by Dobermann and Fairhurst (2000) were replaced where possible with locally relevant values. Nutrient budgets for N and P under the double rice system were calculated for the scenario where 2/3 of the straw is removed, which is current practice, and for 100 % retention of straw as an indication of the consequences of different straw management strategies. Results Preliminary investigation of farmers’ use of waste water The preliminary study showed that rice yields in farmers’ fields using wastewater from fishponds for irrigation were higher than in paddies using an equivalent volume and time of application of river water for irrigation. Yield difference between the two methods was about 1 t/ha (Table 3). This indicates that wastewater, applied at appropriate rates, can help to further increase rice yield. Analysis of soil samples at harvest time showed that total nitrogen, phosphorus and potassium in paddies with wastewater application were significantly higher than plots without wastewater application but organic carbon was lower (Table 4). Wastewater was rich in nitrogen, phosphorus, and potassium (Table 2) which is likely why soils receiving it had higher nutrient contents. By contrast, the high bacterial loading in waste water may accelerate decomposition of organic matter leaving lower organic C levels but higher mineralized nitrogen levels. The survey also recognized that farmers usually added zeolite, lime and dolomite while cleaning fishponds after harvesting (Bosma et al. 2009). This may be the reason for higher contents of calcium and magnesium in paddies receiving wastewater. Besides that, iron and manganese were also significantly higher in wastewater-treated fields (Table 4). Recycling of wastewater for rice cultivation at Chau Phu At Chau Phu, rice yields of all treatments were in the range 3.9- 4.2 t/ha in the wet season 2007 and not statistically different. However, in the dry season 2008 rice yields of T1 and T2 were significantly higher than the other treatments (T3, T4 and T5) with no added P or N or only 33 % of the recommended N (Table 5). This suggests that irrigation by wastewater from fishponds can save 1/3 of recommended N, P and K. The lower yields in T3 were attributed to the acidity of soils in which phosphorus supply is a key factor for crop growth (Cong et al. 1995). Besides that, nitrogen in T3, T4 and T5 was only 33 % of the recommended rate and not sufficient to achieve potential yields for the dry season. Rice yield in the wet season is usually lower than in dry season in the Cuu Long Delta due to lower solar radiation (Hung et al., 1995) Analysis of soil, straw and grain samples at harvesting time showed no significant difference among treatments in concentrations of N, P and K (data not shown). Recycling of wastewater for rice cultivation at Phu Tan At Phu Tan site 1 in the dry season, rice yields in T1 and T2 were close to 7 t/ha and not significantly different (Table 5). This suggests that irrigation by wastewater from fishponds can save 1/3 of recommended N and ½ of the recommended P and K. Further decrease in N fertiliser resulted in reducing yield. Omission of P produced the lowest yield at Phu Tan 1, because low P on these soils limits N use efficiency (Cong et al., 1995). At Phu Tan 2, the maximum yield was only 5.7 t/ha, but T2 again produced the same yield as the recommended fertiliser rate. Further decreases in N and P decreased yield significantly. Macro and secondary nutrient uptake at the Phu Tan sites (Table 6) showed that plots with high yield were also high in nutrient uptake (kg/ha) in straw and grain apart from P in straw of Phu Tan 1. In the experiment at Phu Tan 2, nutrient uptake in grain followed the same trend as in the experiment Phu Tan 1 but K and Ca uptake in straw were not statistically different among treatments (Tables 6). Recycling of wastewater for rice cultivation at CLRRI. At CLRRI, rice yields of all treatment were not different (Table 7). Even though T1, with the highest applied inorganic N, P and K fertiliser rates (100-18-25 kg/ha), had the highest yield, due to high variability among the plots, the effect was not significant. Nitrogen contents in grain at harvesting time of treatments T1, T2 and T4 were among the highest and they were statistically different to others (Table 8). This suggests that differences in N on plots irrigated with waste water from fishpond were not a decisive factor for rice growing on acid soil where P is deficient. Low N content in T3 plot might have resulted from low P application in this treatment as compared to T4. As regard to P content in grain of this experiment, treatment T1 had lower P concentration than others which may result from a dilution effect because this treatment had the highest yield. Recycling of wastewater for rice cultivation at Chau Phu in 2009 At the two Chau Phu sites in the wet and dry seasons of 2009, respectively, there were no significant yield différences among treatments. Even when 40 % of the recommended N and 67 % of the recommended P were applied as fertiliser, no decrease in yield was obtained in the rice irrigated with waste water (Table 7). At Chau Phu, the crops were a healthy green colour during growth with 40 kg N/ha. This suggests moderate but not excessive levels of N in waste water at Chau Phu, perhaps because there is greater use of settling ponds here. Recycling of wastewater for rice cultivation at Phong Dien Total fertiliser application in treatment T1 (irrigated with waste water from fishpond) at planting was about 1/3 of treatment T2 (which used only river water). Later on fertilisers used in treatment T1 was based on plant diagnosis for N (leaf colour), P (tillering capacity) and K (leaf turgour). Overall, T1 had about 45 % less N and 40 % less P application as fertiliser, but a 76 % increase in K in order to minimise insect damage. There was no difference in yields between two treatments over two succeeding crops in 2009 (Table 15). This demonstrated that 40-45 % decrease in N and P fertilisers for rice irrigated with waste water from Clarias fishpond did not induce any nutrient disorders (Tables 16,17). Nutrient budgets Nutrients budgets of 4 sites were presented in Tables 19 and 20. If rice straw was removed, all treatments at Phu Tan and Phong Dien had negative N balances regardless of differences in N dosages of treatments. But experiments in Chau Phu and CLRRI showed that only treatment T5 & T6, with 30 kg N/ha applied as fertiliser had negative N balances. When straw was recycled in situ, N balance was positive for most cases except treatment T5 in Chau Phu, T6 in Phu Tan and T1 in Phong Dien which all had a small N deficit. Phosphorus was in surplus even when straw was completely removed at all 4 sites except treatment T1 at Phong Dien. However, P was in surplus to the extent of 18-70 kg P/ha in the waste water irrigated rice crops if rice straw was retained except if no P fertiliser at all was applied (Table 20). Changes in soil properties [...]... applicable in both wet and dry seasons With the exception of the preliminary yield assessment of farmers’ rice crops irrigated with waste water in 2007, the effect of the waste water irrigation was to reduce fertiliser inputs without affecting yield Hence the main economic benefit of using the waste water would be to decrease fertiliser costs, rather than to achieve increased yield When the catfish farmer... irrigated with wastewater, in most cases the savings to farmers from reduced fertiliser inputs would be the main incentive for use of wastewater The use of fishpond wastewater to irrigate rice can consistently save 1/3 or more of N fertiliser, and up to ½ of the P and K currently applied to crops as inorganic fertiliser Savings could be valued at about 1.16 million VND /ha No phytotoxicity to rice plants... irrigation is small compared to rice and may not make a significant contribution to recycling waste water Alternative treatments need to be considered for the management of wastewater During the wet season, there needs to be a means of treating wastewater other than irrigating rice fields because water levels in the fields will already be high due to heavy rainfall Settling ponds, with or without aquatic plants... need to negotiate with rice farmers for disposal of their waste water If water discharge regulations were enforced, catfish farmers would have a strong incentive to negotiate such arrangements In the absence of regulatory inducements, the onus seems to be on convincing rice farmers of the value of the waste water resource, so that they seek arrangements with catfish farmers for access to waste water. .. affords the opportunity for farmers to greatly reduce their fertiliser costs Indeed, unless farmers reduce N fertiliser rate, they risk decreasing rice yield when applying wastewater due to increased risk of crop lodging Such experiences have been reported to discourage farmers in the past from irrigating with wastewater On the other hand, some An Giang farmers had, through trial and error, reduced their... produces rice, as is common in An Giang Province (Cao et al 2009), the logistics for the transfer of wastewater to the padi field and for the capture of the reduced fertiliser costs within the family business enterprise seems quite obvious and simple By contrast, in many parts of the Mekong Delta the fish farmers run a specialised business with no rice production (Cao et al 2009) Here the catfish farmers... working relationship between catfish farm operators and rice farmers would be necessary so that the timing of wastewater release can be coordinated with irrigation schedules for rice Catfish farmers may seek to recover some of their costs by a payment for the use of the waste water but would obviously need to set the price at less than the cost of the equivalent fertiliser saved by rice farmers Wastewater... millions VND The other benefit from using wastewater to irrigate crops is the cash saving from reduced fertiliser rates In the scenario where fertiliser was reduced to 33 % of N, 50 % for P and for K, the saving in fertiliser costs was equivalent to 1,161 million VND /ha If full straw removal is practiced, reducing fertiliser requirements to 50 % of recommended N, 33 % of P and 0 % of K, the saving in fertiliser... 350,000 ha of rice land may be required to fully treat wastewater in the Mekong Delta Long term effects on soils The long term effects of wastewater application on soil properties are unclear Clogging of soil pores can be a consequence of using wastewater containing suspended particulates However, in padi rice, slowing of percolation rates by pore clogging is not likely to harm crop growth If wastewater was... 15) The main consequence of excess N is lodging of rice crops which may decrease yield Decreasing N fertiliser additions to 33 % of the recommended rate and relying on wastewater irrigation for the remaining N requirements would still generally supply more total N than is required by rice Straw removal would diminish the surplus of N carried forward to the following crop, but in the case of 33 % of the . Technical Report Irrigating rice crops with waste water to reduce environmental pollution from catfish production in the Mekong Delta Cao van Phung 1 , Nguyen be Phuc 2 ,. 1 t of rice is 4 millions VND. The other benefit from using wastewater to irrigate crops is the cash saving from reduced fertiliser rates. In the scenario where fertiliser was reduced to 33. yield when applying wastewater due to increased risk of crop lodging. Such experiences have been reported to discourage farmers in the past from irrigating with wastewater. On the other hand, some