Aquaculture International

, Volume 21, Issue 6, pp 1333–1342

Assessment of AquaMats for removing ammonia in intensive commercial Pacific white shrimp Litopenaeus vannamei aquaculture systems


  • Zhitao Huang
    • Department of FisheriesOcean University of China
  • Rong Wan
    • Department of FisheriesOcean University of China
    • Department of FisheriesOcean University of China
  • Eric Hallerman
    • Department of Fish and Wildlife ConservationVirginia Polytechnic Institute and State University

DOI: 10.1007/s10499-013-9636-7

Cite this article as:
Huang, Z., Wan, R., Song, X. et al. Aquacult Int (2013) 21: 1333. doi:10.1007/s10499-013-9636-7


AquaMats are high surface–area polymer filters whose use produces higher yields with reduced health risks for the aquaculture product. We used AquaMats in pilot-scale systems and in intensive commercial Pacific white shrimp Litopenaeus vannamei production systems to stabilize and improve water quality by removing ammonia. In the pilot-scale systems, evaluation of the effects of temperature and hydraulic retention time (HRT) on ammonia removal rate indicated that the surface total ammonia nitrogen (TAN) conversion rate (STR, mg TAN/m2-day) increased with increasing temperature and decreasing HRT. The highest STR of 319.8 mg TAN/m2-day was observed at a temperature of 30 °C and a HRT of 5 min. In the commercial shrimp production systems, ammonia levels were significantly greater in the control systems (without AquaMats) than in the treatment systems (with AquaMats) after 6 days (P < 0.05). Results suggested that eight 150 cm × 90 cm pieces of AquaMats (0.057 m2 surface area per m3 culture volume) were sufficient for promoting nitrification in this system. The growth rate of juvenile shrimp was most enhanced in treatment C (with 12 pieces of AquaMats, 0.085 m2/m3), which exhibited a significant decrease in ammonia.


AquaMatsAmmonia removalShrimp cultureLitopenaeus vannamei


In aquaculture systems, ammonia wastes originate from animal wastes and from decomposing organic solids such as uneaten feed (Stewart et al. 2006; Kuhn et al. 2010). A large variety of biofilters are used to remove ammonia and other metabolic waste products from recirculating aquaculture systems. Nitrification occurs within biofilters when two distinct groups of autotrophic bacteria catabolize unionized ammonia to nitrite and mineralize nitrite to nitrate (Hagopian and Riley 1998; Chen et al. 2006; Davidson et al. 2008). Unionized ammonia (a portion of total ammonia nitrogen, TAN) and nitrite (which is in acid–base equilibrium with nitrous acid) are toxic to fish and shellfish at relatively low concentrations. Therefore, effective treatment of these compounds is crucial within recirculating aquaculture systems.

In this study, AquaMats fixed-film biofilters (Meridian Aquatic Technology, Calverton, MD, USA) were used to treat dissolved wastes in pilot-scale laboratory systems and in commercial Pacific white shrimp Litopenaeus vannamei culture systems. AquaMats are an artificial polymer substrate with large surface area that is designed to encourage colonization and growth of algae, zooplankton, bacteria, and other aquatic organisms (Arndt et al. 2002). The high effective surface area, 200 m2 per linear meter of material (Ennis and Bilawa 2000), enhances removal of soluble pollutants, including nitrogen, phosphorus, and organic material, from shellfish production units. AquaMats have been used to provide structure in fish culture ponds (Scott and McNeil 2001) and enhance biological processes in ornamental ponds (Ennis and Bilawa 2000). Water with reduced pollutant concentrations can, in turn, enhance aquaculture yield at moderate cost and low energy consumption (Lezama-Cervantes and Paniagua-Michel 2010). Arndt et al. (2002) and Bender et al. (2004) demonstrated that microbial mats and AquaMats efficiently remove ammonia from fish pond effluents, while Erler et al. (2004) quantified significant total suspended solids and nutrient removal using AquaMats in combination with omnivorous fish to treat shrimp farm effluent, given a considerable retention time. Arndt et al. (2002) also showed a transitory beneficial impact on rainbow trout fin length when AquaMats were used. AquaMats also have been used to enhance the growth and survival rate in pilot-scale L. vannamei culture systems (Bratvold and Browdy 2001; Moss and Moss 2004; Otoshi et al. 2006; Audelo-Naranjo et al. 2012).

The objective of this research was to evaluate the performance of biofiltration with AquaMats media in pilot systems and in commercial-scale shrimp production systems. The biofilter was tested in both contexts defined by Colt et al. (2006): (1) on a pilot-scale, allowing evaluation of media under controlled conditions using synthetic wastewater; and (2) on a production system-scale, allowing evaluation of media under production conditions using wastewater from a juvenile shrimp culture system.

Materials and methods

Pilot-scale study

This study investigated the effects of temperature and hydraulic retention time on the biofilter’s ability to remove ammonia from a flow-through system. Four levels of each factor were evaluated, and three replicates were performed for each factor combination. The biofilters were operated at temperatures of 21, 24, 27, or 30 °C and 5, 10, 15, or 20 min of hydraulic retention time.

Pilot-scale systems

The pilot-scale systems were located in the Aquaculture Building, Ocean University of China, Qingdao, China. The systems contained 48 rectangular aquaria (40 cm × 20 cm × 20 cm), into which five pieces of AquaMats (15 cm long × 2.5 cm high each, Model 25004, Meridian Aquatic Technology) were suspended from the water surface, extending to the bottom as a biofilter. The systems also contained a solution reservoir, a pump, oxygen diffuser, and heater (Fig. 1).
Fig. 1

Schematic drawing of experimental installation of AquaMats for pilot-scale studies

Experimental design

The AquaMats biofilters were inoculated uniformly with activated sludge before the study was begun. The activated sludge was obtained from Yellow Sea Fisheries Research Institute hatchery. Flushing the biofilters washed most of the inoculum out of the aquaria. Seawater was obtained from the Yellow Sea near Qingdao. Synthetic wastewater was supplied to the system in order to maintain desired ammonia and nitrite concentrations by way of inlets to the aquaria. The wastewater was a mixture of basal nutrient solution which (per 1.8 m3) was composed of NH4Cl 13.9 g, NaHCO3 35 g, Na2HPO4 1.59 g, KH2PO4 1.53 g, and MnSO4·7H2O 0.36 g (Zhu and Chen 1999). Synthetic wastewater containing approximately 2 mg/L ammonia was fed continuously into the solution reservoir by a metering pump. Continuous feeding was maintained until the outlet TAN concentrations were stable. Experimental measurements were then begun. During the five-week experiment, dissolved oxygen (DO) concentrations were always maintained above 6 mg/L in all the aquaria using oxygen diffusers.

Water quality I

Dissolved oxygen (DO) and pH were monitored daily using a YSI 85 probe (Yellow Springs, Inc., Yellow Springs, Ohio, USA) and a pH meter. Water was sampled every third day for the duration of the experiment. Sample bottles were rinsed three times with water from the sample ports before being filled. Total ammonia–nitrogen (TAN) and nitrite–nitrogen (NO2–N) were analyzed using the phenate and sulfanilamide-NED methods, respectively (APHA et al. 1998).

Surface TAN conversion rate

The surface TAN conversion rate (STR, mg TAN/m2-day) is defined as the amount of ammonia removed by biomass on a unit surface area of medium per day. STR (Colt et al. 2006) was calculated as:
$$STR = 1 , 4 40\;Q\left( {TAN_{\text{in}} - TAN_{\text{out}} } \right)/A \times 100\;\% $$
where TANin and TANout were the influent and effluent TAN concentrations, respectively; Q was the flow rate of the biofilter (L/min); and A was the effective surface area of the AquaMats (m2).

Commercial-scale study

Experimental design

The experimental design for the commercial-scale shrimp production study was a randomized complete block design with four levels of the independent variable (numbers of pieces of AquaMats, Model 25004), each combination with three replicates. Three control systems (A1–A3) did not contain AquaMats biofilters. Treatment B (B1–B3), treatment C (C1–C3), and treatment D (D1–D3) were fitted with 8 (0.057 m2/m3), 12 (0.085 m2/m3), or 16 (0.113 m2/m3) pieces of AquaMats (150 cm × 90 cm each), respectively.

System design

Pacific white shrimp, L. vannamei, were reared in twelve independent round concrete tanks (18 m inside diameter × 1.5 m deep) over a 22-day period at Luyuan Aquaculture, Inc., Zhoushan, China. The AquaMats were suspended from the water surface to the bottom of each production tank. Each tank system (Fig. 2) included a pond aerator. Settleable solids and uneaten feed were drained from the bottom-center drain once a day, and expelled seawater was replaced. The daily water exchange rate was approximately 3.5 %.
Fig. 2

Schematic drawing of side- and top-views of installation of AquaMats in commercial-scale production systems for juvenile Litopenaeus vannamei

Shrimp production

Shrimp were randomly divided into the tanks, resulting in a stocking density of 150 shrimp/m2. Mean initial shrimp weight was 6.4 ± 0.3 g (mean ± standard error) for the control and treatment tanks. Shrimp were fed three times a day, at 9:00, 13:00, and 17:00, at a rate of 7 % body weight per day. Feeding rate was such that excess feed should always be available (Song 2003). Feeding rates were adjusted weekly based on an estimated growth rate of 1 g/week (Kuhn et al. 2010).

Water quality II

Dissolved oxygen (DO), temperature, and pH were monitored twice a day using a YSI 85 probe (Yellow Springs, Inc.) and a pH meter. Water from the tanks was sampled at 9:00 a.m. on every third day. Total ammonia–nitrogen (TAN) and nitrite–nitrogen (NO2–N) were analyzed using the phenate and sulfanilamide-NED methods, respectively (APHA et al. 1998). Salinity was measured using a salinity refractometer (Atago Co., Ltd., Tokyo, Japan).


Statistical analysis was performed using JMP 9 for Windows. Differences between systems were considered significant when P < 0.05.


Pilot-scale experiment

Effects of temperature and hydraulic retention time on ammonia removal by AquaMats biofilters

The STR values for each combination of temperature and hydraulic retention time are shown in Table 1. The ammonia removal rate increased with temperature and decreased with HRT. In all cases, larger mean STR values were attained at higher temperatures. The results indicated that the maximum STR was 319.8 mg TAN/m2-day with an HRT of 5 min and temperature of 30 °C. STR decreased with increasing hydraulic retention time using submerged AquaMats filters. Kim et al. (2000) found similar results when HRTs were 0.6, 3.4, or 16.0 h, the highest ammonia removal rate was 82 g/m3 per day, and when the HRT was at 0.6 h, lowest among all the HRTs. Wortman and Wheaton (1991) found that ammonia removal rate was linearly related to temperature in the range 7–35 °C. Tseng and Wu (2004) indicated that higher removal rate was attained at higher temperature based on the ammonia removal model. Zhu and Chen (2002) found that a temperature increment at 20 °C resulted in a nitrification rate increase of 1.1 % per °C and 4.3 % per °C under DO- and TAN-limited conditions, respectively.
Table 1

Surface TAN conversion rate (mean ± SE) of AquaMats biofilters under various temperature and HRT conditions

HRT (min)

STR (mg TAN/m2-day)

21 °C

24 °C

27 °C

30 °C


291.9 ± 31.6a1

296.1 ± 17.1a1

299.9 ± 8.1a1

303.4 ± 7.1a1


292.9 ± 11.2a1

297.2 ± 10.0a1

301.5 ± 13.3a1

305.5 ± 8.6a1


292.0 ± 15.1a1

297.2 ± 11.0a12

303.1 ± 11.1a12

307.6 ± 14.4a2


298.4 ± 10.8a1

310.3 ± 8.4b12

319.7 ± 9.8b2

319.8 ± 13.8b2

Different superscript letters indicate significant differences among HRT values; different superscript numbers indicate significant differences among temperatures (P < 0.05)

Commercial-scale production experiment

Throughout the experiment, the temperature in the culture water ranged between 27 and 31 °C, and the salinity was 3.4 ± 0.1 g/100 g. Average pH value was significantly lower in the control than in the treatment systems (Table 2). Lowest dissolved oxygen levels were recorded in treatments B and C. We explain these observations by noting that algae attached to the AquaMats can take up carbon dioxide, thereby increasing pH and reducing dissolved oxygen levels.
Table 2

Water quality parameters (mean ± SE) in commercial-scale systems


pH (n = 44)

Dissolved oxygen (mg/L) (n = 44)

Control A

7.89 ± 0.12a

4.76 ± 0.15a

Treatment B

8.09 ± 0.12b

4.44 ± 0.07b

Treatment C

8.04 ± 0.14b

4.62 ± 0.07ab

Treatment D

8.04 ± 0.09b

4.77 ± 0.09a

Different superscripts indicate significant differences among values in columns (P < 0.05)

Mean total ammonia–nitrogen levels decreased from the beginning of the experiment in all the systems, except for the 18th day in the control (A) and treatment D systems (Fig. 3). We attribute the ammonia rise at the 18th day to the rise in shrimp feeding rate from the 15th day. Mean total ammonia–nitrogen did not exceed 0.6 mg/L after the first week in the treatment systems, but did exceed 0.8 mg/L for all 21 days in the control system. Ammonia levels were significantly greater in the control systems than in the treatment systems after 6 days (P < 0.05). It is likely that this outcome was because the AquaMats enhanced nitrification, thereby removing dissolved ammonia. No significant differences were observed between treatments B, C, and D (P < 0.05) except at day 12 (P > 0.05). This finding suggests that eight pieces of AquaMats(0.057 m2 surface area per m3 culture volume) were sufficient to improve the nitrification rate in this intensive shrimp production system.
Fig. 3

Mean total ammonia–nitrogen concentration in commercial-scale shrimp production systems: (a) without AquaMats, or with (b) eight, (c) 12, or (d) 16 pieces of AquaMats

Mean nitrite–nitrogen (Fig. 4) remained around 0.55 mg/L in the control system; in contrast, mean nitrite–nitrogen decreased after the sixth day in the treatment systems, and nitrite levels were significantly (P < 0.05) higher in the control systems (A) than in treatment C after 6 days. L. vannamei can tolerate high concentrations of nitrite (>2 mg/L) (Boyd and Gautier 2000; Crab et al. 2007), but very high levels may seriously inhibit shrimp food intake and growth or prove lethal. We found that the nitrite level was suitably low for the shrimp in all the systems.
Fig. 4

Mean nitrite concentration in commercial-scale shrimp production systems: (a) without AquaMats, or with (b) eight, (c) 12, or (d) 16 pieces of AquaMats

The AquaMats positively affected shrimp growth (Table 3). Percent increases in weight were 46.7, 53.6, 74.1, and 66.7 %, in systems A, B, C, and D, respectively. Feed conversion ratios were significantly lower in treatment systems B, C, and D than in control system A. Shrimp growth rates were significantly greater in treatment systems C and D than in control system A. The final shrimp weights in treatment B, C, and D were 5.4, 19.5, and 15.3 % greater than in control system A, respectively. These differences are smaller than those presented by Bratvold and Browdy (2001) and Moss and Moss (2004), who indicated that final weight of L. vannamei was 28–37 and 34.5 % greater in tanks with AquaMats than without, respectively. The differences in results among the studies may be due to the later life stage, smaller density of AquaMats, and shorter time period in our study.
Table 3

Mean (mean ± SE) initial weights, final weights, feed conversion ratios, and growth rates of shrimp in different treatments


Initial weight (g)

Final weight (g)

FCR (%)

Growth rate (% increase in weight)


6.4 ± 0.3

9.4 ± 3.1a

1.86 ± 0.03a

46.7 ± 5.4a


6.4 ± 0.3

9.8 ± 2.4ab

1.54 ± 0.04b

53.6 ± 4.3ab


6.4 ± 0.3

11.1 ± 1.5c

1.12 ± 0.03c

74.1 ± 6.4c


6.4 ± 0.3

10.7 ± 2.4b

1.37 ± 0.15b

66.7 ± 5.9b

Different superscripts indicate significant differences among values within columns (P < 0.05)

Control systems (A) did not contain AquaMats biofilters, and treatment B included 8 (0.057 m2/m3), treatment C 12 (0.085 m2/m3), and treatment D 16 (0.113 m2/m3) pieces of AquaMats


Different microbial processes may act simultaneously in aquaculture systems, contributing to nitrogen conversion in the mats during bioremediation of the shrimp culture water. For example, the activities of photoautotrophic (algal-based), autotrophic, and heterotrophic bacteria may converge to affect water quality parameters, most importantly ammonia, nitrite, and nitrate levels (Lezama-Cervantes and Paniagua-Michel 2010). Many aquaculture systems include units designed to promote nitrification in order to remove ammonia and nitrite from the culture water. Our research suggests that AquaMats, used in pilot-scale systems and in commercial intensive juvenile shrimp production systems, stabilized and improved water quality by removing ammonia and converting nitrite to nitrate. Hence, use of this filtration material provided an effective tool for achieving high yields at reduced risk to production.

Our work builds upon the results of earlier studies finding utility of artificial substrates for the production of shrimp at various life stages. Bratvold and Browdy (2001) showed increased growth of post-larval L. vannamei in 3.35-m diameter polyethylene tanks using both sand sediment and AquaMats, attributing the increase to the quantity of vertical surface areas in conjunction with improved water quality. Rearing post-larval shrimp in 230-L tanks, Moss and Moss (2004) showed stocking density and substrate effects on final weight, but not an effect on water quality; their results indicate that the AquaMats substrate mitigated the negative effects of high stocking density by increasing the availability of particulate organic matter as a food source for the shrimp. In a 2 × 2 factorial design, Otoshi et al. (2006) grew juvenile shrimp in flow-through 110-L tanks in well water or in water pumped from an intensive shrimp culture pond and either containing AquaMats or not; significant increases in growth were observed using pond water and substrate, although there was no significant interaction effect. Focusing on the impact of AquaMats upon water quality in pilot-scale (1 m3) intensive shrimp production systems, Audelo-Naranjo et al. (2010) showed increased uptake of nitrogen in shrimp biomass and reduced nitrogen discharge in effluents. Within these systems, shrimp production was 13 % higher in systems containing AquaMats, which Audelo-Naranjo et al. (2012) attributed to lower ammonium and ammonia concentrations, and heightened the availability and diversity of food items in the form of periphyton. Using a related species, Arnold et al. (2006) intensively grew post-larval tiger shrimp Peneaus monodon in tank systems at two densities with or without AquaMats and showed that the additional substrate enhanced survival, biomass, and feed conversion ratio. Against this background, we build upon earlier work by showing the effect of temperature and hydraulic retention time on nitrification rate and by demonstrating the utility of AquaMats under full-scale shrimp production under intensive commercial on-growing conditions. We demonstrate that AquaMats contributed to growth rate of juvenile L. vannamei through improvement of water quality in terms of reduced total ammonia nitrogen. Our results suggest that deployment of AquaMats generates significant increases in production at modest cost. In the system we studied, the costs of the AquaMats were $86.40, $129.60, and $172.80 for treatments B, C, and D, respectively. Focusing on treatment C, the yield was approximately 65 kg more than in the control system. Assuming that the price of the shrimp was $3.50/kg, $227.50 in additional value was produced, realizing an increment in profit of $97.90 relative to the control system. We regard this as but a crude estimate of marginal profit, as we believe that use of the AquaMats will shorten production time, an issue that our data do not directly address. We suggest that a complete evaluation of investment cost and production gains would be a suitable topic for near-term research.

Although we did not make observations of feeding behavior, AquaMats can provide additional surface area for shrimp access and grazing, offering an additional food resource supplementing exogenous feed. The particulate organic matter that becomes attached to the AquaMats serves as the food resource for L. vannamei (Moss 2002). Although AquaMats can create operational problems by slowing flow and mixing rates, creating anaerobic areas for sediment to collect, and posing complexities to harvest, farmers still can take advantage of their benefits to improve profitability.


The authors gratefully acknowledge that funding for this study was provided by the National Key Technology Research and Development Program of China (Grant No. 2011BAD13B04). They also thank Luyuan Aquaculture for providing access to facilities for the research. The authors also thank Timothy Lane from Virginia Tech for assistance in the preparation of this manuscript.

Copyright information

© Springer Science+Business Media Dordrecht 2013