Alternative Challenges to Water Quality Improvement and Conservation of Freshwater Bivalves

Abstract

The response surface method has been used by researchers to optimize the conditions of the organic filtration reaction using shellfish. They analyzed the data by setting the size, flow rate, and filtration rate of shellfish as independent variables using the central composite design program in the software. The highest filtration rate was obtained at a flow rate of 24 L/h and a shell size of 10.6–11.4 cm with a residence time of 22.7 hr., while the lowest filtration rate was obtained at a flow rate of 48 L/h with a residence time of 10.4 hr. The excessively fast and low flow rate decreased the filtration rate owing to the increase in the residence time. The researchers suggest that for validation purposes, water velocities in a further range, along with an increase in retention time, should be assessed. They used a central composite design to optimize Sinanodonta woodiana filtration rates and feces production to identify the key factors and optimal conditions. The proposed response surface model illustrated the interactions among the variables on mussel filtration rate. The experimental filtration rate of 4.47 ± 1.82 L/mussel/h corresponded to the predicted value of 8.4 L/mussel/h, which validated the practicability of this optimization strategy.

1. Introduction

Freshwater bivalve species and populations are declining globally due to multiple factors, such as habitat destruction, pollution, loss of host fish for larval development, and invasive species outbreaks [1, 2]. Stream ecologists aim to improve the water quality and conserve mussel species diversity, particularly in the context of climate change. To contribute to these efforts, our laboratory and field studies aimed to understand the ecological role of mussels, including their filtration rates, feces and pseudo-feces production, mortality, impact on water quality, and interactions with other biota [3, 4, 5]. However, our findings indicate that uncontrolled introduction of mussels to field conditions is not an effective method for improving water quality or conserving species diversity, as it may result in increased ammonia tolerance and mortality.

Freshwater unionid mussels, such as Anodonta and Unio species, play a vital role in freshwater ecosystems through the biodeposition of feces and pseudo-feces [6], filter feeding on seston, and linking pelagic and benthic food webs [7, 8]. And, the growth and reproduction of juvenile and adult mussels are influenced by the type and depth of stream sediment as well as the availability of host fish for larval development [9, 10].

In temperate regions, the nutrient content in water systems can quickly increase, leading to the growth of cyanobacteria and diatoms, causing water quality problems. Researchers have attempted various eco-friendly physicochemical methods to improve water quality, such as using food chains or nutritional interactions to promote the growth of animal plankton, which are avian predators, and controlling and managing the optimal growth conditions of predators or competitors. However, these methods have drawbacks, such as difficulties in rapid ecosystem transition and negative effects on water resource development and preservation. Biological manipulation for water quality improvement and ecological restoration is challenging, as it requires accurate prediction of the characteristics of the introduced organisms and anticipating expected changes after biological application while considering the risks associated with introducing exotic species.

Bivalve mollusks are recognized for their high biomass in freshwater and marine ecosystems because of their filter-feeding activity. This activity helps to remove phytoplankton and other suspended matter from the water column, leading to improved light penetration and facilitating the growth of macrophytes that provide habitats for other biota [11, 12, 13]. Studies have demonstrated that greater biological richness is associated with a greater abundance of unionid mussels in freshwater systems [14, 15]. Freshwater bivalves, such as zebra mussels, have been recognized as potential biofilters for drinking water treatment has been recognized [16, 17, 18, 19]. The method of utilizing top-down effects through food chains additionally can be employed to indirectly induce the actions of other organisms. Filter-feeding bivalves, with their excellent filtration capabilities, have been extensively researched, primarily in Europe, for their potential application in water quality improvement. The utilization of these bivalves as unique and diverse biological filters holds significant promise for the removal of organic matter in aquatic environments.

However, the optimal conditions for effective biofiltration systems involving bivalves are still under investigation. Response Surface Methodology (RSM) is a statistical technique used to evaluate the effects of different factors and to identify the optimal conditions for desirable responses in experiments [20, 21]. In this study, we aimed to identify the optimal conditions for particulate material removal by the freshwater mussel Sinanodonta woodiana Lea, using RSM. This large mussel species is native to the Amur and Yangtze river basins, but has spread to other regions due to fish farming [22]. Researchers have investigated the effects of shell size, water flow rate, filtration rate (FR), and production of feces/PF on the removal of suspended material from the water column.

In this study, we aimed to identify the optimal conditions for particulate material removal from a water column using freshwater mussels. Specifically, we investigated the relative importance of shell size, water flow rate, filtration rate (FR), and the production of feces and pseudo-feces (PF) using RSM. We focused on the freshwater mussel Sinanodonta woodiana Lea, which is a large species native to the Amur and Yangtze River basins and has spread throughout Southeast Asia and South America due to fish farming. Our study proposes the use of domestic freshwater shellfish to control algae in highly polluted agricultural reservoirs, establishing a sustainable and ecologically friendly water improvement method using response surface analysis. We used a central composite design to efficiently create approximate secondary or higher response functions.

2. Methods and materials

2.1 Test animal collection

We collected adult Anodonta woodiana Lea and Unio Douglasiae Griffith et Pidgeon from tributaries of the Geumgang and Mangyeonggang Rivers to serve as spawning sites for fish. Shellfish are commonly found in domestic water systems and usually coexist in the same river; however, Anodonta woodiana Lea is abundant in organic matter, and Unio Douglasiae Griffith et Pidgeon is relatively common in the upper stream of the river in irrigation channels with developed sand or pearls. They primarily feed on plankton or bottom organic matter [22]. We immediately placed the shellfish collected in an icebox and transported them to the laboratory. After washing them two to three times with desalinated tap water, they were kept in a management tank for more than three months. We acclimated S. woodiana specimens in laboratory aquaria for at least three months which were collected directly from waterways and streams in Korea. We used stainless steel treatment baths (80 × 80 × 145 cm) with a working volume of 500 L to study filtration. We acclimated 30 individuals of test mussels to holding aquaria for 18 days before the commencement of the experiments. The experimental equipment used has been described in detail by Lee et al. [23]. An experiment was conducted involving filtration by S. woodiana under conditions including a water temperature of 19 ± 3°C, water flow rates ranging from 24 L/h (T1) to 48 L/h (T2), and a photoperiod of 12 D:12 L.

2.2 Removal of organic matter by freshwater bivalves

We measured the filtration rate and production of bivalve feces and pseudo-feces using a device with the same structure as Lee et al. [3] using the continuous organic material removal method with shellfish. The device consisted of a storage tank, treatment tank, and analysis tank. We made a treatment tank of stainless steel (80 × 80 × 145 cm) with a total treatment capacity of 500 L. We installed a lattice-type shellfish holder (75 × 75 cm) under 10 cm of the check surface in the treatment tank and placed the shellfish on it. The surface water of the eutrophic reservoir was passed through the tank at 24 and 48 L/h. The water temperature was adjusted to 18 ± 2°C and the optical instrument was maintained at 12:12 L. We supplemented shellfish of a similar size to those that died during the experiment. To understand the effect of shellfish treatment on water quality, we measured changes in the water quality environment daily and analyzed the samples collected from the analysis tank to observe changes in organic and nutritional salts. We also collected shellfish excrement every five days to measure the production of excrement.

We collected data on the filtration rates for each animal every 42 days (Table 1). The production of feces and pseudo-feces by each mussel was also measured simultaneously, with sediments from individual mussels being analyzed at three-day intervals for nine days during the 42-day experiment. The ash-free dry mass of each of the 30 mussels was measured at the end of the experimental period.

Characters	Operation conditions
Kind of mussels	Anodonta woodiana
Mussel density (mussel/L)	0.5
Total weight (kg)	10.5–30.1
Shell length (cm)	8.09–11.35
Shell width (cm)	4.74–6.71
Total AFDW (g)	1.33–3.42
Tank volume (L)	500
Flow rate (L/h)	22.0
Total capacity (㎥/d)	0.52
Filtration area (㎡)	0.64
Temperature (°C)	11.9–21.8
Exposure time (days)	42
Sampling time (every)	AM 11:00

Table 1.

The experimental design and operation conditions for the removal of organic matter by the freshwater bivalve Anodonta woodiana.

We harvested sedimented feces and pseudo-feces from the experimental aquaria and placed them in sterilized dishes. The weight of the feces was measured after drying at 70°C for 1 h. The production of mussel feces and pseudo-feces was quantified by calculating the disparity in dry weights (mg/g AFDM/h) of sedimented particulate matter between reactor treatments with and without mussels, employing the subsequent equation: PFs were determined using the equation: PFs = (V/M) × ln(T/C)/t, where V represents the volume of the experimental chamber in liters, M signifies the total ash-free dry mass (AFDM) of the mussels, T and C denote the total dry weights of sedimented particulate matter in the reactor with and without mussels, respectively, and t stands for the duration of the experiment in hours. After detaching the entire mussel body from the shell and measuring its weight, the mussel was placed into a heat-resistant container. Subsequently, it was subjected to desiccation at 100°C for 20 minutes in a drying oven until a consistent mass was achieved. Following this, the mussel was incinerated in a muffle furnace at 500°C for 2 hours, as per the methodology outlined in APHA 1995. The ash-free dry mass (AFDM) of the mussel body was calculated by utilizing the disparity in dry weight before and after the incineration process. The filtration rate of each mussel (FR: L/mussel/h) was determined for each experiment and day using the equation established by Coughlan [24]: FR = (V/M) × ln(T/C)/t, where V represents the volume of the experimental reactor in liters, M signifies the total ash-free dry mass (AFDM) of the mussels, T and C denote the concentrations of suspended solids in water that passed through the reactor with and without mussels, respectively, and t (hours) stands for the duration of the experiment.

2.3 Experimental design and the modeling of filtration by mussels

We analyzed the data to optimize the organic filtration reaction conditions using shellfish by applying statistical methods and the response surface method (RSM) through Minitab software (MINITAB Release 14.12.1, Korea). We considered the size, flow rate, and filtration rate of shellfish as independent variables and utilized the central composite design (CCD) program in the software to set their high and low levels. The size of the shellfish was set to 2 cm and 20 cm for high and low levels, respectively, while the flow rate was set to 48 L/h and 24 L/h for high and low levels, respectively. Table 2 presents the experimental design used in this study, including the trends, interactions, and experimental errors of the operating variables.

Variables	Range of levels
	Actual	Coded	Actual	Coded	Actual	Coded
Mussel size (cm)	5.0	−1	8.0	0	12.0	+1
Water flow (L/h)	12.0	−1	24.0	0	48.0	+1
Filtration rate (L/mussel/h)	0.5	−1	1.0	0	2.0	+1

Table 2.

Experimental range and levels of the three independent variables used in response surface methodology in terms of actual and coded factors.

An experimental design was employed to determine the optimal levels of three variables—mussel size (×1), experimental time (×2), and water flow (×3)—regarding their effects on filtration rates and feces production. This design aimed to model mussel filtration conditions with respect to body size (Table 3). Using 23 central composite designs (CCD) with three factors, including six center points, a series of 30 experiments were conducted, with each experimental condition involving a single individual mussel. All variables were standardized to a central coded value of zero. The ranges encompassing the minimum and maximum values of the investigated variables, along with the comprehensive experimental plan detailing their actual and coded forms, are provided in Table 4.

	Shell length (cm)	Flow rate (L/h)	Retention time (h)	Filtration rate (L/mussel/h) (mean ± SE)
Sinanodonta woodiana	8.5 ± 1.0	12	1.5	0.87 ± 0.17
Sinanodonta woodiana	11.4 ± 1.8	24	22.7	4.47 ± 1.82
Sinanodonta woodiana	11.4 ± 1.3	24	3.9	1.23 ± 0.20
Sinanodonta woodiana	10.6 ± 1.9	24	20.8	3.30 ± 1.03
Sinanodonta woodiana	10.8 ± 1.9	48	10.4	2.30 ± 0.97

Table 3.

Filtration rate of the mussel Anodonta woodiana according to differences in flow rate and retention time.

Run order	Shell size (×1)	Flow rate (×2)	Production of feces (×3)	Mean measured response	Predicted response
1	−1	−1	−1	2.36	3.18
2	−1	−1	1	3.40	4.59
3	−1	−1	−1	3.57	4.81
4	−1	−1	1	2.75	3.71
5	−1	1	−1	2.77	3.73
6	−1	1	1	4.06	5.48
7	−1	1	−1	3.21	4.32
8	−1	1	1	1.72	2.33
9	1	−1	−1	2.58	3.47
10	1	−1	1	2.62	3.53
11	1	−1	−1	6.13	8.27
12	1	−1	1	4.78	6.45
13	1	1	−1	2.81	3.79
14	1	1	1	2.47	3.33
15	1	1	−1	13.45	18.14
16	1	1	1	10.71	14.45
17	1	0	0	1.69	2.28
18	1	0	0	6.15	8.29
19	0	−1	0	1.43	1.93
20	0	1	0	4.78	6.45
21	0	0	0	3.97	5.35
22	0	0	0	2.38	3.21
23	0	0	−1	6.53	8.80
24	0	0	1	6.09	8.21
25	0	0	0	6.11	8.24
26	0	0	0	6.45	8.70
27	0	0	0	6.55	8.83
28	0	0	0	6.53	8.80
29	0	0	0	6.28	8.48
30	0	0	0	6.08	8.20

Table 4.

Experimental designs used in response surface methodology using three independent variables with the centre point showing measured and predicted values of the filtration rates of freshwater bivalve Anodonta woodiana.

After conducting the experiments, The average maximum filtration rate was employed as the dependent variable or response (Y), and a second-order polynomial equation was applied to the dataset through the utilization of the multiple regression procedure.

This led to the development of an empirical model that establishes a correlation between the responses and the independent variables of the experiment. The formulated model equation for the three-factor system is expressed as follows: Y = β0 + β1 x1 + β2 x2 + β3 x23 + β11 x12 + β22 x22 + β33 x32 + β12 × x1 × x2 + β13 × x1 × x3 + β23 × x2 × x3, where Y represents the predicted response, β0 denotes the intercept, β1, β2, and β3 stand for the linear coefficients, β11, β22, and β33 represent the squared coefficients, and β12, β13, and β23 signify the interaction coefficients. The statistical analysis of the data was conducted using the Minitab software package (Minitab Release 14.12.1).

2.4 Data analysis

We used a HORIBA U-22XD (HORIBA Ltd., Japan) to measure the temperature, pH, electrical conductivity, dissolved oxygen (DO), and turbidity of both the control and treatment groups daily at 11:00 AM. We collected samples from the analysis tank at the same time each day to observe the changes in organic and nutritional salts. We filtered the Chl-a sample with a GF/F filter (Whatman Inc., England), extracted it with 90% acetone for 24 h, and separated it by centrifugation for 20 min following the Water Quality Process Test Act [25]. For nutritional was measured NO₂-N by the Colorimetric method, NO₃-N using the cadmium reduction method, NH₃-N using the phenate method, TN using the cadmium reduction method, PO₄-P using the ascorbic acid method, and TP using the ascorbic acid method after persulfate decomposition [26]. We also monitored the water temperature and light intensity at 30-minute intervals during the experiment using a HOBO Pendant Temperature/Light Data Logger (UA-002-08). Light intensity was used as an indirect indicator to observe changes in the turbidity of organic matter in the water. We conducted ANOVA and Tukey’s THD test using the SPSS package (version 12.0.1, SPSS Inc., 2004 release) to determine the correlation between the application of shellfish and changes in the water quality according to the flow rate; the significance level was set at P < 0.05 (Table 5).

Parameters	Unit	Control	T1 (24 l/h)	T2 (48 l/h)	F	P
Temperature	°C	16.09 ± 3.41	15.75 ± 3.22	15.79 ± 3.19	0.03	P = 0.971
Conductivity	μS/cm	66.12 ± 88.83	67.66 ± 90.12	67.71 ± 89.36	0.001	P = 0.999
DO	mg/L	11.83 ± 2.53b	8.36 ± 1.88a	6.92 ± 1.54a	13.87	P < 0.001
Turbidity	NTU	26.37 ± 10.14b	8.85 ± 4.53a	6.84 ± 3.35a	23.13	P < 0.001
pH	—	8.77 ± 0.24	8.54 ± 0.19	8.53 ± 0.33	2.23	P = 0.128
SS	mg/L	22.58 ± 3.17b	5.96 ± 1.20a	3.93 ± 2.40a	163.27	P < 0.001
Chl-a	μg/L	24.77 ± 8.99b	7.42 ± 4.67a	3.98 ± 2.68a	30.50	P < 0.001
NO2-N	μg/L	4.10 ± 2.05	8.68 ± 6.85	8.27 ± 2.43	3.03	P = 0.067
NO3-N	mg/L	0.12 ± 0.04	0.17 ± 0.03	0.17 ± 0.04	3.67	P = 0.041
NH4-N	μg/L	22.29 ± 23.98a	261.88 ± 94.86b	472.20 ± 92.38c	75.55	P < 0.001
TN	mg/L	1.49 ± 0.23b	1.30 ± 0.12a	1.67 ± 0.14b	9.70	P < 0.001
PO4-P	μg/L	7.45 ± 2.09a	8.57 ± 4.39a	49.05 ± 22.57b	28.44	P < 0.001
TP	μg/L	98.83 ± 16.40b	58.83 ± 8.58a	115.41 ± 24.37b	24.38	P < 0.001

Table 5.

Summary of ANOVA on environmental quality by the stocking water of freshwater bivalve Anodonta woodiana.

Control; only lake water without mussels, DO; dissolved oxygen, SS; suspended solids, Chl-a; chlorophyll-a, TN; Total nitrogen; and TP; total phosphorus, Letters (a, b and c) indicate significant differences by ANOVA and Tukey’s THD test.

3. Results and discussion

3.1 Effect of mussel size, flow rate, and retention time on Sinanodonta filtration rate

Table 2 shows that the mussel treatments caused changes in several water quality parameters. As expected, turbidity, suspended sediments, and algae (as reflected by chl a) were significantly reduced by filtration. The excretion of feces and pseudo-feces by mussels was caused in the significantly higher concentrations of ammonium and orthophosphate as phosphorus in those treatments. Across various mussel sizes (mean value ± SE: 8.5 ± 1.0 to 11.4 ± 1.8 cm), flow rates (12–48 L/h), and retention times (1.5–22.7 h), the filtration rate exhibited a range spanning from 0.87 ± 0.17 to 4.47 ± 1.82 L/mussel/h (with an average of 2.67 ± 1.00 L/mussel/h; see Table 4). Filtration rate displayed an increment with the enlargement of mussel size and a decrease in flow rate. Specifically, the group of larger mussels (11.4 ± 1.8 cm) exhibited higher filtration rates in comparison to the smaller size group (8.5 ± 1.0 cm).

3.2 Optimization by response surface methodology

Response surface methodology was used to optimize the filtration rate of Sinanodonta woodiana in response to three independent variables: mussel size, water flow rate, and retention time. Table 4 presents the results of the CCD experiments and the mean predicted and observed responses. After analyzing the variance (ANOVA), the researchers obtained an R2 value of 0.7625, indicating that the quadratic model had a satisfactory adjustment to the experimental data and could explain 76% of the variability in the response. The coefficients of the regression equation was obtained: Y = 18.214−10.211x1 + 10.105x2 + 12.542x3−12.458x12−8.243x22−9.549x32 + 13.263x1 x2 + 17.671x1 x3−15.842 × 2 × 3.

We determined the interaction between the experimental components and the optimum of each component required for the maximum filtration rate by using three-dimensional response surface curves. Figures 1 and 2 illustrate the comparative impacts of two variables, namely mussel size and flow rate, across different retention times. The researchers anticipated the mussels’ highest attainable filtration rate to reach 8.4 L/mussel/h within these parameters, aligning with the maximum levels (+1) of mussel size (13.0 ± 0.2 cm) and flow rate (17.5 L/h; depicted in Figure 1). Nevertheless, the curve also illustrates the responsiveness altering in concurrence with the water flow velocity. As the flow rate escalates (surpassing 30 L/h) and the retention time diminishes, the feces production by mussels escalates further, reaching 11.1 g AFDM/ind./h (as depicted in Figure 2). The curves of the response surface exhibited a lack of curvature, adopting a flattened disposition. While mussel size exhibited a relatively marginal influence, higher flow rates were associated with amplified fecal production (as seen in Figure 2). Figure 3 provides insight into the higher fecal production rates linked to elevated flow rates, while indicating a limited impact of mussel size on fecal production.

Figure 1.

Response surface curve of filtration rates by S. woodiana showing interaction between mussel size and water flow.

Figure 2.

Response surface contour plots of filtration rates by S. woodiana showing interaction between mussel size and water flow.

Figure 3.

Response surface contour plots of feces production by S. woodiana showing interaction between mussel size and water flow.

We recognized the necessity for optimizing conditions, including the size and filtration rate of organisms, particularly in eutrophic regions, when employing freshwater mussels for organic matter removal. The surface plots indicate a requirement for reducing the flow rate and extending the retention time to minimize fecal production. In light of these findings, researchers propose the implementation of a further reduction in water flow and an extension of retention time within the system as a means of validation. The researchers also suggested that a multifactorial analytical approach, which takes into account the interaction of independent variables, provides a basis for models designed to assess the nonlinear nature of the response under limited experimental conditions.

3.3 Nutrient changes

The shellfish treatment group caused a significant change in the concentration of nutrient salts depending on the flow rate, resulting in a higher value of NO2-N (C:0.03 ± 0.01, FL − 1:0.11 ± 0.01, FL-2:0.07 ± 0.01LL-1) and NH3-N (C:36.77 ± 6.87, FL-1:306.09 ± 16.31, FL-2:214.99 ± 14.57LL-1) compared to the high flow rate treatment group (FL-2) in the low flow rate treatment group (FL-1). Shellfish emit ammonia-type nitrogen during the eating process, and the amount of nitrogen emitted during the eating process is higher than that of phosphorus. Other experiments with various organisms have also shown an increase in underwater dissolved inorganic nutrients owing to shellfish treatment, which calls for further studies on the removal, concentration, and recovery of the generated nutritional salts for use by other organisms or plants. Increasing the transparency of water using shellfish creates a favorable environment for the growth of submerged plants or attached algae, thereby converting the basis of the food chain from plant plankton to attached algae or submerged plants. Fukushima et al. [27], Kim Deung [22, 28, 29], and Lee [23] conducted experiments that found that shellfish treatment increased underwater dissolved inorganic nutrients. Future studies are needed to determine how to remove, concentrate, and recover the generated nutritional salts so that they can be used by other organisms or plants.

3.4 Filtration rates

The shellfish treatment group had a higher average filtration rate (CR) than the control group (2.299 ± 0.97 L mussel−1d−1) at a flow rate of 48 L/h in the experiment to determine the effect of shellfish on organic material control according to flow rate. However, Kirby-Smith [30] and Jorgensen et al. [31] reported that shellfish’s ability to filter particulate matter in water is reduced at high flow rates, and this study similarly showed low CR values at fast flow rates. The highest filtration rate was observed at a flow rate of 24 L/h and a shell size of 10.6–11.4 cm with a residence time of 22.7 hr., while the lowest filtration rate was observed at a flow rate of 48 L h − 1 with a residence time of 10.4 hr. The excessively high flow rate decreased the filtration rate, and the excessively low flow rate also decreased the filtration rate owing to the increase in residence time. Lee et al. [3] reported a lower filtration rate (0.87 ± 0.17 L /mussel/d) at a flow rate of 12 L/h, which was believed to be due to the long residence time affecting the reproduction of plant plankton. Previous studies have shown that the presence of shellfish can increase the growth of plankton. An optimal flow rate range for each shellfish exists, and for the continuous organic material control device using pearl shells, the optimal flow rate was found to be 24 L/h. Future studies can reveal the relationship between the flow rate and residence time and improve device efficiency.

3.5 Response surface analysis

We used the reaction surface analysis method to determine the optimal conditions for controlling organic material in shellfish. We determined the size and flow rate of the shellfish-dependent variables and conducted a multi-session and dispersion analysis using software. We also performed dynamic analysis and three-dimensional analysis. To design the experimental conditions, we used the central synthesis design method, setting the size of the shellfish to a high level of 2 cm, a low level of 20 cm, a flow rate of 48 L/h, and a low level of 24 L/h. The results in Table 4 were obtained by performing the organic matter control reaction of shellfish, with the organic filtration rate as the dependent variable.

The results showed that the shellfish organic material control rate was higher when the size of the shellfish was larger and the flow rate was lower. We conducted an analysis in which we fixed the size (10 cm) and flow rate (36 L/h) as the independent variables. This analysis showed that the larger the size of the shellfish and the slower the flow rate, the higher the organic material control rate by the shellfish. However, environmental factors such as water temperature can affect the ability of shellfish to control irregular organic material. We also derived a contour plot of the relationship between the response factor and the response. Individual contour plots were obtained by fixing the size and flow rate of the shellfish to normal values. The results showed that both the size and flow rate of the shellfish are important factors in the organic material control reaction, and that stabilizing external environmental factors, such as water temperature, is also crucial for achieving stable organic material control.

Therefore, to increase organic material control using shellfish in a stable manner, we should consider not only the size, flow rate, and residence time of the shellfish but also the stabilization method of external environmental factors that affect the activity of shellfish.

3.6 Potential applications

Ismail et al. [32] highlighted that the utilization of environmentally relevant concentrations and treated wastewater can offer an initial insight into the potential effectiveness of bivalves in mitigating emerging contaminants and enhancing water quality. The application of the response surface methodology approach revealed optimal conditions, which could vary based on factors such as season, bivalve species, age, and prey competition, thereby necessitating additional research. While existing models reported in relation to eutrophication reduction using bivalves within lake systems can offer insights, their direct applicability for on-site organic matter removal is limited due to the absence of correlations between mussel filtration rates and rates of algae and organic matter removal.

Nonetheless, investigations have substantiated the noteworthy filtration efficiency and the utilization of bivalves for substantial water quality enhancement, thereby signifying the potential viability of bivalves in engineered systems or ecological rehabilitation efforts [19, 33]. The choice and upkeep of an apt bivalve species and population, alongside the fine-tuning of bivalve filtration rates and feces production, emerge as pivotal factors when evaluating the suitability of bivalve application for enhancing water environments, as advocated by these researchers.

The proliferation of dense algal blooms and the accumulation of excessive organic materials within the water column can degrade the habitat quality for various organisms and severely disrupt essential ecosystem services, including recreational activities, sanitation, irrigation, industrial cooling, and the supply of potable water. Consequently, numerous nations have endeavored to mitigate algal blooms and enhance the quality of eutrophic water bodies. Nonetheless, a significant drawback lies in the potential harm posed by various physical and chemical treatment approaches, which can trigger adverse environmental consequences, including secondary pollution. In response, the focus of this study is directed toward the management of organic matter levels and the population density of problematic algae, employing readily adaptable technologies. The native freshwater mussel S. woodiana possesses distinct advantages, owing to its robust resilience in challenging environments, including instances of harmful algal blooms (HABs). Notably, it can thrive at notably elevated densities, reaching levels as high as 60 individuals per square meter [34]. Additionally, with a lifespan of up to 12 years [35], this species proves conducive to environmental management strategies that require minimal maintenance.

From a pragmatic standpoint, our study’s outcomes propose that the passage of nutrient-enriched water through mussel beds could establish effective biological filters. The optimal elimination of suspended sediments is attainable through the careful selection of appropriate mussel sizes, flow rates, and retention times. These systems have the potential to function as artificial conduits, seamlessly connecting to lentic or lotic water bodies. Furthermore, this technology lends itself to integration with other sustainable approaches for nutrient management, including the introduction of phytoplanktivorous fish [36] and the cultivation of rapidly proliferating macrophytes [37].

4. Conclusion

Using this response surface method, we optimized the conditions for the organic filtration reaction of shellfish. We analyzed the data using the central composite design program in the software, with the size, flow rate, and filtration rate of shellfish as independent variables.

• The experiment showed that the highest filtration rate was achieved by increasing the flow rate to 24 L/h and using shell sizes of 10.6–11.4 cm, with a residence time of 22.7 h. Conversely, the lowest filtration rate was observed with a flow rate of 48 L/h and a residence time of 10.4 h. Both excessively fast and excessively slow flow rates decreased the filtration rate, attributed to factors such as residence time. However, the model could only be validated with mussel size, and further validation is required with a broader range of water velocities and retention times.

• The utilization of response surface methodologies necessitates a limited number of individuals for the determination of optimal filtration parameters, rendering them a cost-effective resource for water management. The presented model equation vividly depicts the quantitative impact of variables and their interactions on mussel filtration rates. The conducted experiment validated the feasibility of the optimization strategy within the optimum conditions of mussel size (13.0 ± 0.2 cm) and flow rate (17.5 L/h), resulting in an observed filtration rate of 4.47 ± 1.82 L/mussel/h, aligning closely with the predicted value of 8.4 L/mussel/h.

From the above results, excessively fast or slow water flow induces a decrease in plankton filtration rates by bivlaves, and it was evident that such filtration rate reduction significantly depends on the residence time within the apparatus. Through further research, elucidating the relationship between water velocity and residence time, and incorporating it into device design, it is anticipated that a higher filtration efficiency can be achieved.

Thus, our research revealed that statistical design methodology is an effective and feasible approach to promote high filtration and low fecal production, indicating that indigenous bivalves offer significant potential to restore multiple ecosystem services provided by freshwater systems by removing organic matter from eutrophic systems.