Optimization of Algicidal Activity for Alteromonas sp. FDHY-03 against Harmful Dinoflagellate Prorocentrum donghaiense

Abstract

Prorocentrum donghaiense is a harmful-algal-bloom-forming species of planktonic dinoflagellates widely distributed around the world, which threatens the marine environment and human health. Bacteria are promising biological agents to control algal growth in HABs. Previously, we isolated an Alteromonas sp. FDHY-03, a P. donghaiense-lysing bacteria strain, from Xiapu Sea area of China. In order to improve the algicidal activity of this strain, we optimized the medium composition and culture conditions. Based on single-factor method experiment design, the optimum medium component of algicidal effect for strain FDHY-03 was sucrose and peptone. The result of Plackett–Burman design indicated that three significant factors (sucrose, peptone, and rotational speed) appeared. Finally, the concentrations of key factors were confirmed by central composite design (CCD) and response surface methodology (RSM). Under the optimized medium, the algicidal rate of strain FDHY-03 against P. donghaiense improved by 67.15%, and the OD₆₀₀ value increased by 2.86-fold. The optimal source and condition were sucrose 0.46% (w/v), peptone 4.25% (w/v) addition, and rotation speed 255 rpm. Overall, this study provides an optimized method and optimum medium for improving the algicidal activity against P. donghaiense, and has a positive influence on algae-lysing bacteria for controlling the blooms of the algae in the environment.

1. Introduction

Red tide is used as a conversational term for one of the natural phenomena known as harmful algal blooms (HABs) that occur in estuarine, marine, or fresh water [1]. HABs can produce harmful or toxic substances which cause a variety of adverse effects, including the destruction of food chains, economic losses, and even endangerment of human health [2,3]. In recent decades, the global spread of harmful algae and aggravation of the hazards have attracted the attention of some relevant international organizations [4]. Up to now, there have been three main methods to control HABs: physical, chemical, and biological methods. Because of its low toxicity, highly efficient, and environmentally friendly characteristics, especially algicidal bacteria [5], the biological method is superior to the other two approaches in controlling HABs [6]. Algicidal bacteria present two strategies to lyse harmful algae cells: one is to directly attack algae cells, and the other is indirect algae-lysing via releasing algicidal compounds [7]. Previous studies have reported that the optimization of culture conditions and medium components of some bacteria has successfully improved the algicidal activity, e.g., Pseudoalteromonas SP48 against Alexandrium tamarense [8], Bacillus sp. strain S51107 exhibits strong algicidal activity against Microcystis aeruginosa [9], and Loktanella sp. Gb-03 inhibits Dinoflagellates [10]. It is necessary to optimize culture objectives to enhance bacteria algicidal activity and cell density when applied to controlling algae bloom.

Studies show that the growth and activity of microorganisms are influenced by the composition of the substrate (C:N ratio, inorganic salts) and growth conditions such as temperature, pH, and oxygen [11,12]. In addition, optimization of the culture medium can affect the growth of algicidal bacteria and the algicidal effect. At present, there are many ways to optimize the selection of medium, including one-factor design, orthogonal design, uniform design, and response surface methodology (RSM). In the process of exploring the optimal culture medium, experimental design and statistical methods play important roles [13]. For example, Li et al. [14] optimized the nutrient source of Mangrovimonas yunxiaonensis LY01 by single-factor design, which improved the growth of bacteria and its algicidal activity against Alexandrium tamarense. Based on single-factor investigation and the uniform design method, the largest dry weight of strain DH46 was obtained and the algicidal rate increased nearly 10% after optimization [13]. Three minerals (KNO₃, MnSO₄·H₂O, and K₂HPO₄) were identified as significant factors that obtained the highest cell concentration of the algicidal bacterium Enterobacter sp. NP23 [15]. The traditional ‘‘one-factor-at-a-time” approach disregards the complex interactions among various physicochemical parameters [16,17]. In addition, this method is time- and cost-consuming, needing larger consumption of reagents and materials [18]. Response surface methodology (RSM) is a useful statistical technique commonly used for the optimization of multivariable systems. It uses quantitative data in experimental design to determine and simultaneously solve multivariate equations in order to optimize processes or products [19,20]. This method has been successfully applied in many areas such as biological, chemical, and pharmaceutical production. Previously, the algicidal Alteromonas sp. FDHY-03, which could produce algicidal compounds against Prorocentrum donghaiense, was isolated from the Xiapu Sea area of China [21]. Members of the genus Alteromonas are widely distributed residents in the marine environment. It is one of the algae-lysing bacteria that has been found [22]. For instance, Alteromonas sp. A14 was applied to red tide of dinoflagellate, Cochlodinium polykrikoides [23]. Alteromonas sp. KNS-16 produced four active algicidal compounds to inhibit algae, such as Heterosigma akashiwo, Cochlodinium polykrikoides, and Alexandrium tamarense [24]. However, there is not any previous related work using RSM methods for the optimization of Alteromonas in order to control algae bloom and optimize algicidal bacteria against P. donghaiense.

Prorocentrum donghaiense is a well-known red-tide-forming marine dinoflagellate widely distributed in China, especially in the East China Sea [25,26]. Previous studies indicated that P. donghaiense is a kind of non-toxic dinoflagellate, which can cover a large marine area and maintain high density and high biomass for over one month [27,28]. During the bloom progress, the color of water changes into reddish-brown, lasting for several days to months, which results in great losses in economy and leads to severe damage to the marine environment [29]. The purpose of this study was to seek a method to increase bacteria density and algicidal active substances for algicidal bacteria strain FDHY-03. The single-factor design was used to optimize the medium components that had an influence on algicidal compounds production, and the fermentation conditions or significant factors were also optimized using Plackett–Burman design and central composite design (CCD) for strain FDHY-03 to produce algicidal compounds.

2. Materials and Method

2.1. Algal Cultures, Algicidal Bacterium Culture, and Seed Bacteria Culture

P. donghaiense was obtained by the Center for Collections of Marine Algae in Xiamen University (strain CCMAXU-364). The algae for the experiments were cultured in fresh sterilized f/2 medium (without silicate) with 0.22 µm filtered nature seawater (29 ± 1 PSU) at 20 ± 1 °C and the light condition was set to 14 h: 10 h light–dark cycle with the light intensity of 100 μmol photons m⁻² s⁻¹. The algae culture was treated with three antibiotics before the following experiment (ampicillin, kanamycin, and streptomycin were 200 mg/L, 100 mg/L, and 100 mg/L, respectively) to minimize the growth of bacteria [21].

Alteromonas sp. FDHY-03 was isolated from the Xiapu Sea area of China, which could produce algicidal compounds to remove P. donghaiense [21]. Strain FDHY-03 was stored at −80 °C in ZoBell 2216E medium with 50% (v/v) glycerol. The 2216E medium for culturing the bacterium FDHY-03 contains 5 g/L of peptone, 1 g/L of yeast extract, 0.01 g/L of FePO₄, and 30 g/L of sea salt, and the final pH was adjusted to 7.2 ± 0.1.

Bacterial colonies were inoculated into Erlenmeyer flasks (100 mL) containing 35 mL ZoBell 2216E medium with shaking at 150 rpm for 24 h at 25 °C to obtain seed bacteria culture.

2.2. Measurement of Algicidal Rate and Biomass

To study the algicidal rate of FDHY-03 on P. donghaiense, 1% volume (v/v) of the bacterial culture were added into 10 mL experiment algae for 48 h, and the intensity of the algal cell was monitored at the end of each experiment. P. donghaiense cells were fixed with Lugol’s solution. Cell density was counted using a Sedgwick-Rafter counting chamber under a microscope. In the control group, the same volume of sterile ZoBell 2216E medium was added to the target algal cultures. All experiments were performed with three replicates.

Algicidal rate was calculated according to the following equation:

$$\mathrm{algicidal}\mathrm{rate}=\left(N_c-N_t\right)/N_c$$

(1)

where N_c represents the number of algal cells in the control group, and N_t represents the number of algal cells in the treatment group.

The bacteria density was determined by measuring the absorbance at 600 nm (OD₆₀₀) with an ultraviolet spectrophotometer (UV-2600, Shimadzu, Kyoto, Japan).

2.3. Single-Factor Design

In order to select the most beneficial carbon source and nitrogen source for the growth of bacteria and production of algicidal compounds for target algae, we used the basal medium (2216E excluded peptone in the whole single-factor design), and the optimization of medium components are as follows: (1) sucrose, soluble starch, glucose, molasses, mannitol, and lactose were used as different carbon sources and added to the basal medium; (2) peptone, yeast extract, urea, soy flour, corn powder, and KNO₃ were used as different nitrogen sources and added to the basal medium. A total of 0.5% (w/v) of different carbon sources and nitrogen sources were added into basal medium. A total of 1% volume (v/v) of the seed bacteria culture were inoculated into 35 mL 2216E broth at 25 °C for 24 h at 150 rpm to obtain fermentation broth of strain FDHY-03. All samples were used to measure cell density and algicidal rate. Different treatment groups were set, and all experiments were performed in triplicate.

2.4. Plackett–Burman Experimental Design

In order to optimize the medium components and culture on bacterium FDHY-03, eight different independent variables were selected as experimental factors. The Plackett–Burman design was applied. Each variable has two levels, high (+1) and low (−1). The results of Plackett–Burman experimental design were fitted by the first-order model as follows:

$$R={\beta }_0+{{\displaystyle \sum }}^{}{\beta }_i{x}_i(\mathrm{i}=1,2,3\dots \dots )$$

(2)

where R represents algicidal rate, ${\beta }_0$ represents the model intercept, $x$ is the regression coefficient, and $x_i$ is the independent factor.

Each variable was defined at two levels, as shown in

Table 1. Different level of variables in Plackett–Burman design.

Variables	Low Level	High Level
	−1	+1
A Sucrose (%, w/v)	0.5	1.5
B Peptone (%, w/v)	1	1.5
C Medium volume (mL)	60	100
D Rotation speed (rpm)	100	200
E Temperature (°C)	20	30
F pH	6	9
G Fermentation time (h)	24	36
H Inoculum amount (%, v/v)	1	2

. Plackett–Burman experimental design runs for 12 times and algicidal rate are presented in

Table 2. Twelve-runs Plackett–Burman design matrix of eight variables and algicidal rate of strain FDHY-03.

Run	A Sucrose	B Bacterial Peptone	C Medium Volume	D Rotation Speed	E Temperature	F pH	G Fermentation Time	H Inoculum Amount	Algicidal Rate (%)
1	+1	−1	+1	+1	+1	−1	−1	−1	35.5948
2	+1	+1	−1	+1	+1	+1	+1	−1	56.7844
3	+1	+1	+1	−1	−1	−1	−1	−1	19.9814
4	−1	+1	+1	+1	−1	−1	−1	+1	56.2268
5	+1	+1	−1	−1	−1	+1	−1	+1	44.5167
6	−1	+1	−1	+1	+1	−1	+1	+1	59.2937
7	+1	−1	−1	−1	+1	−1	+1	+1	25.1394
8	+1	−1	+1	+1	−1	+1	+1	+1	32.2491
9	−1	−1	−1	+1	−1	+1	+1	−1	45.4926
10	−1	−1	−1	−1	−1	−1	−1	−1	39.777
11	−1	−1	+1	−1	+1	+1	−1	+1	30.7156
12	−1	+1	+1	−1	+1	+1	+1	−1	40.8922

.

2.5. Central Composite Design

Central composite design (CCD) is widely used in response surface methodology (RSM) to obtain a second-order model. In this study, three significant variables were applied, with 5 levels (−1.68179, −1, 0, +1, +1.68179), as shown in

Table 3. Different level of variables in CCD design.

Variables	−α	Low Level	0	High Level	+α
	−1.68179	−1	0	+1	+1.68179
A Sucrose (%, w/v)	0.06	0.2	0.4	0.6	0.74
B Peptone (%, w/v)	2.24	2.75	3.5	4.25	4.76
D Rotation speed (rpm)	199.55	220	250	280	300.45

. CCD experimental design for 20 runs and algicidal rate are presented in

Table 4. Twenty-runs central composite design matrix of three variables and algicidal rate of strain FDHY-03.

Run	A	B	D	Algicidal Rate (%)
1	0	0	0	96.2307
2	0	0	0	93.0814
3	+1	−1	+1	72.4895
4	0	0	0	93.9241
5	+α	0	0	91.1674
6	0	0	0	95.1617
7	0	−α	0	77.5527
8	+1	−1	−1	89.0295
9	0	+α	0	97.6371
10	−1	+1	−1	88.0169
11	0	0	−α	88.9620
12	0	0	0	91.7300
13	−1	+1	+1	90.9058
14	+1	+1	−1	89.7609
15	−1	−1	−1	86.2785
16	0	0	0	96.4557
17	−1	−1	+1	80.3094
18	−α	0	0	92.3994
19	0	0	+α	83.0661
20	+1	+1	+1	92.4051

. Multiple regression analysis can be applied to predict the dependent variables on the basis of a second-order equation.

$$R={\beta }_0+{{\displaystyle \sum }}_i{\beta }_i{x}_i+{{\displaystyle \sum }}_{ii}{\beta }_{ii}{x}_i^2+{{\displaystyle \sum }}_{ij}{\beta }_{ij}{x}_i{x}_j$$

(3)

where R represents algicidal rate, ${\beta }_0$ represents model intercept, ${\beta }_i$ is the $i$th linear coefficient, ${\beta }_{ii}$ is the $i$th quadratic coefficient, and ${\beta }_{ij}$ is the $ij$th interaction coefficient.

2.6. Statistical Analysis

The data were analyzed by a single-factor analysis of variance (ANOVA) using SPSS version 26. Three replicate samples were expressed as mean ± SD. Bars without the same letter represented significant difference (p < 0.05). The average of each group was ordered from largest to smallest to derive a, b, and c. Design-Expert 8.0.6.1 was used for Plackett–Burman design, CCD, and the regression analysis of the experimental data.

3. Result

3.1. Optimization of Medium Components on Strain FDHY-03

The effect of carbon sources on the algicidal effect from FDHY-03 was evaluated using six different carbon sources in conical flasks, and the results are shown in Figure 1a. After shaking flask fermentation, the algicidal rate of the bacteria was 19.28% when sucrose served as the culture medium’s carbon source, which is more significant than other carbon sources, excluding mannitol. When sucrose was used as the carbon source of the culture medium, the bacterial density (OD₆₀₀ value) was 8.496, and the values were significantly higher than other carbon sources (p < 0.05). Therefore, sucrose was selected as the optimal carbon source in culture. Optimization of nitrogen source and the results are shown in Figure 1b. When peptone and urea were selected as nitrogen sources of culture medium, there was no significant difference in algicidal rate of bacteria after shaking flask fermentation between the two groups, but it was obviously higher than other nitrogen sources. The biomass of FDHY-03 had its maximum value in the medium containing peptone (OD₆₀₀ = 2.25). Moreover, urea had an algaecide effect. Therefore, peptone was selected as the optimal nitrogen source in culture.

3.2. Plackett–Burman Design

The eight selected factors (sucrose, peptone, medium volume, rotation speed, temperature, initial pH, fermentation time, inoculum amount) in Plackett–Burman design and their levels are shown in Table 1. The Plackett–Burman design matrix selected to screen the significant variables along with the algicidal results are shown in Table 2. While the strain was cultured in the sixth run, the algicidal rate of the P. donghaiense achieved the highest values, almost reaching 59.29%. From the results, it can be seen that there is a wide range in algicidal rate (from 19.9814% to 59.29%), which indicates that the optimized experiment has a great influence on algicidal rate.

Shown in

Table 5. ANOVA and significant analysis for Plackett–Burman design.

Source	Sum of Square	df	Mean Square	F Value	p-Value
Model	1698.50	8	212.31	11.85	0.0333
A	281.61	1	281.61	15.72	0.0287
B	393.61	1	393.61	21.97	0.0184
C	255.24	1	255.24	14.25	0.0326
D	596.70	1	596.70	33.31	0.0103
E	8.63	1	8.63	0.48	0.5376
F	17.85	1	17.85	1.00	0.3917
G	137.14	1	137.14	7.65	0.0698
H	7.71	1	7.71	0.43	0.5586

is the first-order response model in the form of analysis of variance (ANOVA). The “F-value” of 11.85 and values of “Prob > F” less than 0.0500 indicate that the model is significant. The effects of A (sucrose), B (peptone), C (medium volume), and D (rotation speed) on the algicidal rate are significant in eight variables (p < 0.05). Both medium volume and rotation speed are related to dissolved oxygen. Among four identified significant variables, the medium volume was excluded because of its low significance. A (sucrose), B (peptone), and D (rotation speed) were determined for further optimization studies. The experimental data were fitted to a first-order polynomial equation, and the effects of various variables on the Plackett–Burman design were determined. The regression equation is as follows:

$$R=41.36-4.84\mathrm{A}+5.73\mathrm{B}-4.61\mathrm{C}+7.05\mathrm{D}+0.85\mathrm{E}+1.22\mathrm{F}-3.38\mathrm{G}+1.60\mathrm{H}$$

(4)

where A, B, C, D, E, F, G, and H represent sucrose, bacterial peptone, medium volume, rotation speed, temperature, initial pH, fermentation time, and inoculum amount, respectively. The regression equation obtained from ANOVA showed that the R-squared was 0.9693. The adjusted determination coefficient (Adj R-squared) was 0.8875 for the model, which indicates that the model is suitable for the Plackett–Burman design.

3.3. Central Composite Design

To search for the optimal value of sucrose, peptone, and rotation speed for algicidal rate, respectively, the central composite design and RSM were applied. The CCD experiment was carried out (altering the direction of three significant variables’ levels) according to the result of the Plackett–Burman experiment. In this study, medium volume, temperature, and pH were fixed at zero level, and fermentation time and inoculum amount were fixed at the lowest level. The design matrix and algicidal results of 20 runs are shown in Table 4. The results showed that the minimum algicidal rate was observed in the third and seventh runs. The algicidal rate of 11 groups were over 90% (1st, 2nd, 4th, 5th, 6th, 9th, 12th, 13th, 16th, 18th, 20th), which indicated that the optimized experiment was effective.

According to the data from the second-order response model, the results of the analysis are shown in

Table 6. ANOVA and significant analysis for central composite design.

Source	Sum of Square	df	Mean Square	F Value	p-Value
Model	780.81	10	78.08	17.18	0.0001
A	1.11	1	1.11	0.24	0.6326
B	326.34	1	326.34	71.82	<0.0001
C	51.18	1	51.18	11.26	0.0084
AB	8.64	1	8.64	1.90	0.2013
AC	14.62	1	14.62	3.22	0.1064
BC	98.30	1	98.30	21.63	0.0012
A2	25.79	1	25.79	5.68	0.0411
B2	114.50	1	114.50	25.20	0.0007
C2	169.09	1	169.09	37.31	0.0002
ABC	13.33	1	13.33	2.93	0.1209
Lack of Fit	23.64	4	5.91	1.71	0.2826
Cor Total	821.71	19

. The “F-value” of 17.18 implies that the model is significant. There is only a 0.01% chance that a “Model F-Value” this large could occur due to noise. The “lack of fit F-value” of 1.71 implies that the lack of fit is not significant relative to the pure error. The high F-value and insignificant lack of fit indicates that the model is a good fit. The determination coefficient R² of 0.9827 indicates that the model was fitting the strain FDHY-03 algicidal rate during the fermentation very well. Moreover, the adjusted determination coefficient (Adj R-squared = 0.8949) indicated that the model has a strong correlation. A low value of the coefficient of variation (CV = 2.39%) demonstrated that the experiments conducted were precise and reliable. To obtain the maximum algicidal rate corresponding to the optimum levels of these significant variables, a second-order regression equation was given, as follows:

$$R=94.49-0.29\mathrm{A}+4.89\mathrm{B}-1.94\mathrm{D}+1.04\mathrm{AB}-1.35\mathrm{AD}+3.51\mathrm{BD}+1.34{\mathrm{A}}^2-2.82{\mathrm{B}}^2-3.43{\mathrm{D}}^2+1.29\mathrm{ABD}$$

(5)

where A, B, and D represent sucrose, bacterial peptone, and rotation speed.

To obtain the optimal levels and interaction of the three variables for maximum algicidal rate, the contour plot and response surface curve plots were obtained by regression equation (Figure 2, Figure 3 and Figure 4). The contour plot can explain the significance of each variable. Results showed that the interaction between rotation and peptone was the most significant one, followed by the rotation and sucrose, and sucrose and peptone. The three-dimensional response surface graph was convex, which indicates that there are three well-defined variables.

The optimum conditions of algicidal rate finally optimized by software were as follows: sucrose 0.46% (w/v), peptone 4.25% (w/v), and rotation speed 256.81 rpm, respectively, and the other variables were kept at zero level and lower level.

3.4. Verification of the model

In order to verify the accuracy and reliability of the model, other conditions remain unchanged on the basis of central composite design; the optimized conditions were adjusted as follows: sucrose 0.46% (w/v), peptone 4.25% (w/v), and rotational speed 255 rpm. An experiment was conducted and compared to the predicted data. Algicidal rate of strain FDHY-03 and the result of bacterial density in different medium results are shown in Figure 5. The algicidal rate of the strain FDHY-03 to P. donghaiense in optimized culture medium reached 98.91%, which is close to the predicted value of 96.84%, showing high accuracy. Furthermore, the OD₆₀₀ value is 12.176 by using the optimized medium; however, after 48 h, the algicidal rate of basic medium (2216E) was 59.17%, and OD₆₀₀ value was 3.152. Compared with that grown in 2216E medium, both algicidal rate and bacteria density of the culture were increased significantly in optimized medium. The results exhibited that the above model was suitable for optimizing the medium in the algicidal rate through two-step optimization by FDHY-03. Bacteria cultivated in the optimized culture medium had higher growth and algicidal rate compared with those cultivated in 2216E medium.

4. Discussion

The growth and dissipation of HABs are often accompanied by changes in bacteria communities [30,31]. It is emphasized that the utilization of certain bacterium can control or mitigate HABs without destroying the environment or other aquatic species [32]. Biology is the most widely used method, especially for its microbial specificity, high efficiency, and environmental friendliness [33]. Therefore, microorganisms have been widely used to control HABs. Since the 1990s, P. donghaiense has formed large-scale HABs in the East China Sea, causing huge economic losses and serious marine environmental damage [34]. Previously, we isolated the algicidal activity of the Alteromonas sp. FDHY-03 against the harmful P. donghaiense [21]. We expected to improve the algicidal efficiency of this algicidal strain in this study.

The growth of bacteria is comprehensively influenced by nutrients, environment, and other factors [35]. Carbon sources and nitrogen sources and their concentrations in the medium play an important role in commencing the production of primary and secondary metabolites, as a limited supply of an essential nutrient can restrict the growth of microbial cells or product formation [36]. Optimizing the culture medium and culture conditions are essential to bacterial fermentation. In this study, we used the traditional one-factor-at-a-time approach to select the optimal carbon source and nitrogen source for algicidal substances. The results indicated that sucrose is an optimal carbon source, which is consistent with Streptomyces sp. HJC-D1 [11]. Addition of glucose was shown to cause the highest algicidal activity of Vibrio sp. co-culture against Akashiwo sanguinea [37]. Furthermore, nitrogen source is essential for the growth of bacteria; it can constitute the protein, nucleic acid, and other nitrogen compounds of organisms [38]. Bacterial algicidal compounds are known to be produced in the presence of preferred organic nitrogen sources. Peptone, the optimum nitrogen source for strain FDHY-03, is consistent with that of Vibrio brasiliensis H115 reported by Li et al. [9]. Yeast extract is the preferred nitrogen source for the marine algicidal bacteria Sphingomonas sp. DC-6 and it obtained the maximum biomass in submerged fermentation [39]. Studies state that bacteria shows an increase in secondary metabolite production utilizing a complex carbon source. On the other hand, concentration of carbon source leads to the change of metabolite levels, which when higher will inhibit the growth of bacteria and the synthesis of its active substance product [40]. We concluded that in the fermentation process, bacteria began to grow sucrose addition, and peptone was the main nutrient involved in the growth and metabolism of bacteria in the middle and late stages. Peptone may be more helpful to produce algicidal compounds for strain FDHY-03, and sucrose acts as the most important factor for bacterial growth.

For different microbial strains, there is no universally defined culture medium. The selection of each method will lead to different culture conditions. Lyu et al. [8] optimized algicidal bacterium Pseudoalteromonas sp. SP48 by orthogonal design and uniform design, and the result was that the biomass of bacteria increased by 79.2% and its lethal dose 50% (LD 50) decreased by 54.0%. Shu [41] reported Talaromyces purpurogenus YL13, and algicidal activity was improved with response surface methodology. RSM overcomes the disadvantages of single-factor tests because it is fast, reliable, and helpful in understanding the influence of the interactions between various factors, thus saving time and manpower [42]. In our study, on the basis of single-factor investigation, the Plackett–Burman design, and the response surface methodology, we screen out significant factors which have the greatest influence on algae-lysing rate. Three significant factors were selected for enhanced algicidal rate using the Plackett–Burman design. The wild range of algicidal rate (from 19.98% to 59.29%) indicates that the experiment needs to be further optimized. The algicidal activity of FDHY-03 increased by 67.15% by using RSM. In addition, the OD₆₀₀ of bacterium increased by 2.86 times from 3.152 to 12.176, which indicated that the algicidal rate and cell density was significantly enhanced after optimization. The 3D plots can directly reflect the effect of different levels of the factors on the response and therefore pinpoint their optimum levels [43]. As a result of our experiments, the optimal fermentation conditions for strain FDHY-03 were determined as follows: source 0.46% (w/v), peptone 4.25% (w/v), and rotation speed 255 rpm. Contour plots can reflect the interaction between two variables. The interaction between A and C shows that peptone and rotation speed have important influence on producing active substances. Alteromonas is a heterotrophic, aerobic, and polarly flagellated bacteria [22,44]. Rotation speed has a direct effect on dissolved oxygen, and proper dissolved oxygen is beneficial to the growth and metabolism of the microorganism.

The close relationship between the predicted and experimental response values from the validation experiment demonstrated the validity and acceptability of the statistical model for the optimization of culture. The results show that RSM is a useful method for the optimization of microbial cultivation. Therefore, it provides a scientific and effective strategy for inhibiting HABs.

5. Conclusions

The medium for fermentation by Alteromonas sp. strain FDHY-03 to produce a strong algicidal effect against P. donghaiense was investigated through two-step optimization. Single-factor design was used to select the optimized carbon and nitrogen sources, and, finally, sucrose and peptone were selected, respectively. From the eight factors of Plackett–Burman design, three significant factors appeared, which were sucrose, peptone, and rotation speed. Based on the results of CCD and RSM, the optimum conditions of these three factors were as follows: sucrose 0.46% (w/v), peptone 4.25% (w/v), and rotation speed 255 rpm. Finally, the algicidal rate of the strain FDHY-03 to P. donghaiense reached 98.91%, which was significantly higher than the 67.15% algicidal rate before optimization. The OD₆₀₀ of bacterium also increased by 2.86-fold. Compared to the basic medium, the optimized culture and culture method in this study is promising in maintaining the population of algicidal bacteria with high biomass and alga-lysing activity. Overall, this study provides a potential optimized strategy and optimum culture medium for improving the algicidal activity of an algicidal bacteria against its target harmful algae at laboratory conditions, and provides a theoretical basis for its application under natural population conditions.