Citrus Juice Marination Improves the Flavor of Fish: A Case Study of Sea Bass (Lateolabrax japonicus)

TL;DR

Marinating sea bass in citrus juices enhances flavor and sensory quality, reducing fishy smell and improving texture.

Contribution

This study demonstrates that citrus juice marinades effectively improve the sensory and flavor properties of sea bass.

Findings

Citrus juice marinades increased volatile compounds like ethyl butyrate and limonene, enhancing flavor.

Orange and grapefruit juice marinades reduced fish hardness and improved sensory scores (p < 0.05).

Lemon juice treatment caused significant whitening of the fish (p < 0.05).

Abstract

Although aquatic products are abundant in premium proteins and other vital nutrients, their unique fishy smell often restricts public acceptance and the development of related products. Therefore, pre-marinating is usually used to improve sensory quality and mitigate fishy smell. In this study, sea bass filets were marinated for 1 h at a solid–liquid ratio of 3:5 (w/w) using 15% orange juice, 15% grapefruit juice, and 10% lemon juice. Subsequently, their effects on the flavor and sensory quality of sea bass were examined. The results of gas chromatography-ion mobility spectrometry (GC-IMS) showed that marinating the filets in citrus juice led to a notable increase in volatile compounds, including esters with fruity flavor such as ethyl butyrate and terpenes with pleasant citrus aromas such as limonene. The results of texture profile analysis (TPA) showed that pre-marinading with orange juice and grapefruit juice significantly reduced the hardness of fish and resulted in higher sensory evaluation scores (p < 0.05). Lemon juice treatment produced the most noticeable whitening effect (p < 0.05). These findings suggest that citrus juices can be effectively incorporated as marinade ingredients for sea bass, providing both theoretical insight and practical guidance for improving the marinating technology for aquatic products.

Figures

Full text

1. Introduction

In recent years, aquatic products have gained significant importance as a vital source of animal protein. Sea bass (Lateolabrax japonicus) is renowned for its tender texture, high protein levels, and low fat, holding significant economic value as a commercially important marine fish species in global fisheries [1]. According to the “China Fishery Statistical Yearbook” [2], the sea bass aquaculture production in China exceeded 240,000 t in 2025, accounting for about 12% of the total marine fish aquaculture output, ranking third among all farmed marine species. Nevertheless, current product forms remain largely confined to fresh sales or frozen whole fish, reflecting the single product form [3]. Consumer acceptance and product development of sea bass are largely constrained by its sensory quality and fishy odor [4]. During cold-chain logistics and pre-cooking handling at home, issues such as fishy smell, oxidative (unpleasant) flavor, or textural hardening can significantly reduce preference.

Common processing methods for sea bass mainly include salting, drying, and smoking. These traditional sea bass products require heavy salting during processing. This often leads to an overly firm texture in the final products, which is increasingly inconsistent with the preferences of modern consumers [5]. There is extensive research [6] highlighting that high salt intake is strongly linked to a greater likelihood of increased hypertension and cardiovascular diseases. In addition to improving sea bass’s texture and palatability, which are two of the most critical characteristics determining consumer preference for products [7], marinating the fish also improves its flavor and reduces its fishy odor. The meat manufacturing industry frequently uses mechanical and chemical techniques to make meat products softer [8]. Çimen et al. [9] pointed out that a variety of additives were employed throughout the marinating process to improve the meat’s color, flavor, and softness. As awareness of healthy eating gradually increases, more and more consumers tend to choose clean-label products marketed as “additive-free”, “preservative-free”, or “natural”, while the acceptance of products containing food additives in the ingredients table decreases [10].

Acidic marinades made from fruits and vegetables have been widely used by researchers in recent investigations. The low pH value of acidic marinades can weaken muscle structure, increasing tenderness [8]. Gök et al. [11] found that pomegranate juice and red grape juice can be used as natural tenderizers for turkey breast meat. Nour [12] found that sour cherry and plum juices improved both the quality and sensory scores of pork loin, alongside improving storage stability. Blueberry, raspberry, and strawberry marinades have been shown in studies [13] to lessen the production of oncogenic heterocyclic amines in grilled beef. Although the effects of fruit juice on the quality and sensory evaluation of poultry meat are well documented, far less is known about its effects on fish, particularly when assessed using an integrated flavor–sensory–quality framework. The fish’s original flavor may be lessened by using juice marinades. Their pleasant smell can enhance the overall flavor of the fish.

Citrus fruits are popular all over the world due to their appealing color, delicious flavor, and wonderful smell [14]. Their juice is rich in antioxidant compounds. Citrus fruits have been found to contain over 170 antioxidants, including phenols, vitamins, and terpenoids [15]. These functional ingredients can provide health benefits beyond basic nutrition. Park et al. [16] demonstrated the beneficial effects of citrus on cancer cells. Citrus fruit juice, which has a unique aroma and a high concentration of organic acids, is acknowledged as a natural flavoring ingredient. Unal et al. demonstrated that the acidity of fruit juice induces swelling of muscle protein, expands and changes light reflection characteristics, thereby changing the color of meat products [17]. Sena et al. noted that an acidic environment can weaken muscle fiber structure and enhance protein degradation, thereby contributing to the effectiveness of citrus juice marinades in tenderizing meat and improving its texture [18]. When used as a marinade, citrus juice can improve fish flavor and lessen its fishy odor. In addition, different citrus juices have differences in characteristic volatile profiles [19], which may lead to different degrees of aroma contribution, resulting in different marinating effects.

Three different kinds of citrus juices were used to marinate sea bass in the current study. Using methods including electronic nose (E-nose), gas chromatography-ion mobility spectrometry (GC-IMS), and electronic tongue (E-tongue), the impact of these citrus juices on the flavor profile of sea bass was evaluated. In addition, by analyzing physical and chemical indicators such as texture profile analysis (TPA), water-holding capacity (WHC), color difference, and microstructure, the impacts of citrus juice on the quality and sensory attributes of sea bass were examined. The study is intended to evaluate the efficacy of citrus juice marinades on the physicochemical properties and flavoring characteristics of sea bass. The research results provide a theoretical and practical foundation for the production of marinated sea bass products.

2. Materials and Methods

2.1. Chemical Reagents

Ethanol, hematoxylin, eosin, Coomassie Brilliant Blue R-250, potassium chloride, tartaric acid, hydrochloric acid, and potassium hydroxide were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All reagents were of analytical grade (purity > 99%).

2.2. Sea Bass Samples and Treatments

A total of 15 fresh iced sea bass (620 ± 10.37 g in weight, 36 ± 1.45 cm in length) were purchased at a nearby seafood market (Zhoushan, China). The sea bass were delivered to the laboratory in a plastic box filled with crushed ice. Oranges, grapefruits, and lemons were bought from a local supermarket (Huarun Wanjia Supermarket, Zhoushan, China). The fish head, skin, and bone were removed, and the back muscles were cut into 4 cm × 1.5 cm × 1.5 cm pieces.

The fish were randomly split into five groups at random (n = 3 fish per group): fresh untreated (F), soaked with pure water (C), marinated with 15% orange juice (O), marinated with 15% grapefruit juice (G), and marinated with 10% lemon juice (L). Marinade solutions consisted only of citrus juice and pure water, and no other ingredients were added. The selected processing parameters (juice concentrations, marination time, and solid–liquid ratio) were determined based on our process optimization trials conducted prior to the present study. A lower concentration was selected for lemon juice due to its higher acidity to avoid over-acidification and undesirable quality changes during marination. Fish samples were marinated for 1 h at a solid–liquid ratio of 3:5 (w/w) under refrigerated conditions at 4 °C. After marination, samples were gently blotted with absorbent paper to remove surface liquid. WHC, TPA, color, and sensory evaluation were conducted on the same day after marination. All remaining samples were packed in sealed bags and stored at −80 °C for subsequent analyses.

2.3. Water-Holding Capacity

The water-holding capacity was determined following the method of Li et al. [20]. About 4 g of the fish sample was weighed and recorded as w1. The samples were wrapped with filter paper and centrifuged at 5000× g for 10 min at 4 °C (MULTIFUGEX3R, Thermo Fisher Scientific Inc., Waltham, MA, USA). The weight of the fish after centrifugation was recorded as w2. The calculation formula is as follows:

$$[eqn]$$

2.4. Texture Profile Analysis

Texture Profile Analysis (TPA) was performed to determine the hardness and springiness of sea bass according to the method of Chen et al. [21] with slight adjustment. The texture analyzer (iTexture, Zheke Instrument Equipment, Hangzhou, China) used a flat-bottom probe P/5 = 5 mm to compress the fish samples (4 cm × 1.5 cm × 1.5 cm) twice. The test conditions were as follows: the compression deformation rate of 30%, trigger force of 5 g, pre-test speed of 1 mm/s, test speed of 1 mm/s, and hold time of 5 s.

2.5. Determination of pH and °Brix

The pH value of fish samples was determined according to the method of Xiao et al. [22], slightly modified. Minced fish samples weighing 5 g were homogenized with 45 mL of deionized water. After standing at 4 °C for 30 min, the supernatant was determined by a pH meter (PHS-3C, INESA Scientific Instrument Co., Ltd., Shanghai, China). The pH meter was calibrated using standard buffer solutions at pH 4.00, 6.86, and 9.18 prior to the measurements.

To provide compositional context for marinade comparability, the diluted citrus juices used for marination (10% lemon juice, 15% orange juice, and 15% grapefruit juice) were measured for acidity and soluble solid content expressed in pH and °Brix, respectively, after dilution. Juice pH was determined using the calibrated pH meter described above. Soluble solid content (°Brix) was measured using a digital refractometer (PAL-1, ATAGO Co., Ltd., Tokyo, Japan). The results are summarized in Table S1.

2.6. Measurement of Color

The values of L*, a*, and b* of the samples were immediately recorded using a handheld colorimeter (DS-200, CHNSpec Technology Co., Ltd., Hangzhou, China). The total color difference was calculated according to the following formula [23]:

$$[eqn]$$

where ΔL**, Δa**, Δb** represent the differences in the L, a*, and b* values of fresh and marinated samples. The color change in fish filets after marinating is more noticeable with the higher ΔE.

2.7. Sensory Evaluation

The sensory evaluation was carried out with the method described by Maxwell et al. [24] and appropriate modifications were made. A trained panel of ten volunteers (20–40 years old; 5 males and 5 females) participated. The assessment was performed on raw samples by visual inspection, odor perception, and tactile assessment (pressing), without tasting or swallowing.

The panelists underwent standardized training tailored to the specifics of this study, and the definitions of descriptors and the scoring scales were calibrated with the aid of reference samples. All evaluations were carried out in a quiet, odor-free room with uniform light. Samples were placed in odor-free containers, labeled with random three-digit codes, and presented blind in a randomized order. Assessors were not allowed to communicate during the session. To avoid olfactory fatigue, panelists were instructed to rest for at least 1 min between consecutive samples.

Each assessor evaluated all coded samples in three independent sessions, with the presentation order re-randomized in each session. Scores from the three rounds were averaged per assessor per sample prior to statistical analysis. Four attributes were scored: color, smell, texture, and elasticity. Each attribute was rated using a structured 20-point scale (1–5, 6–10, 11–15, 16–20) with specific criteria defined in Table S2. The overall sensory score (maximum = 80) was obtained by summing the four attribute scores (equal weighting).

2.8. Microstructure Observation

According to the method of İncili et al. [25], the sample was cut into 1 cm × 1 cm × 1 cm blocks and fixed with 4% paraformaldehyde at 4 °C for 24 h. After fixation, samples were dehydrated through ethanol, embedded in paraffin, and then sliced. Sections were stained with hematoxylin and eosin, dehydrated, cleared, and sealed. Microstructure images were acquired using high-resolution digital scanning equipment (KF-PRO-120, Konfoong Bioinformation Tech Co., Ltd., Ningbo, China).

2.9. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis Analysis

SDS-PAGE was performed using the Yang et al. [26] method with a few adjustments. The concentration of MPs was adjusted to 1 mg/mL, and 5 μL loading buffer was added to the 20 μL sample. The mixture was heated in a water bath at 100 °C for 5 min, and then centrifuged at 5000× g for 5 min. Finally, 10 μL of the resulting supernatant was collected for loading. The power supply voltage was 120 V, running for 1 h. Following the running procedure, the gels were first stained with Coomassie Brilliant Blue R-250 solution, then decolored and imaged.

2.10. Determination of Smell Characteristics by Electronic Nose

The smell of filet samples was analyzed by the PEN3 electronic nose (PEN3, AIRSENSE Company, Schwerin, Germany). The detection system is composed of 10 sensors. The response characteristics of each sensor are as follows: W1C (benzene/aromatic), W3C (ammonia/aromatic), W5C (alkanes/aromatic), W1W (inorganic sulfides), W2W (aromatic/organic sulfides), W1S (short chain alkanes), W2S (alcohols/ethers/aldehydes/ketones), W3S (long chain alkanes), W5S (nitrogen oxides), W6S (hydrides). The samples were prepared according to the method of Wu et al. [27] with minor modifications. The 5 g chopped sample was transferred to a 100 mL beaker and sealed with a double-layer plastic wrap. The sample was tested on the machine after standing for 1 h at 25 °C. Conditions for determination were as follows: sampling time was 1 s/group; the self-cleaning time of the sensor was 80 s, sensor zeroing time was 5 s, the sample preparation time was 5 s, the injection flow rate was set to 400 mL/min, and the analysis sampling duration was 80 s.

2.11. Taste Trait Determination by Electronic Tongue

Taste analysis was conducted using an E-tongue (TS-5000Z, INSENT Company, Kawasaki, Japan). The procedure followed the method described by Wang et al. [28], with slight modifications. For the analysis, 5 g fish samples were weighed and homogenized with 15 mL deionized water for 1 min, then centrifuged (10,000× g, 10 min, 4 °C). Collect the supernatant, filter it twice with filter paper, and analyze the resulting clear filtrate using an electronic tongue taste system. Each sample was measured 4 times in parallel, and the data from the following 3 times were taken for analysis.

2.12. Gas Chromatography-Ion Mobility Spectrometry

A method described by Zhu et al. [29] was adopted for the analysis of volatile compounds among different sample groups, using a GC-IMS (FlavourSpec^®^, GAS Company, Dortmund, Germany) coupled with a MXT-WAX capillary column (30 m × 0.53 mm, 1 μm, Restek Company, Bellefonte, PA, USA). For each analysis, 3 g of the sample was placed in a 20 mL headspace vial and incubated at 60 °C for 15 min before injection. A 500 μL sample was injected using a syringe preheated to 85 °C. Each fish-level sample was analyzed in triplicate, and the mean signal intensities were used for subsequent statistical comparisons. The retention index was calculated based on the retention time of the target compounds. The target compounds were identified by matching the obtained retention index and drift time against the corresponding data in the GC-IMS library.

2.13. Statistical Analysis

Unless otherwise specified, each treatment group included n = 3 fish (biological replicates). Where applicable, each fish-level sample was analyzed in triplicate (technical replicates), and the per-fish mean value was used for statistical analysis. Technical replicates were not considered independent biological replicates and were averaged per fish prior to statistical inference. Data are presented as mean ± standard deviation. Prior to parametric testing, normality and homogeneity of variances were examined using the Shapiro–Wilk test and Levene’s test, respectively. Differences among treatments were assessed by one-way ANOVA, followed by Duncan’s multiple range test for post hoc multiple comparisons. Different letters indicate significant differences among treatments at p < 0.05. Each endpoint was analyzed separately. No formal adjustment was applied for multiplicity across different endpoint families. Given the large number of endpoints and the small biological n, findings should be interpreted as exploratory. All statistical analyses were performed using IBM SPSS Statistics (Version 27.0; SPSS Inc., Chicago, IL, USA). Figures were generated using OriginPro 2024 (OriginLab Inc., Northampton, MA, USA). GC-IMS data were processed and visualized using the Reporter plugin, Laboratory Analytical Viewer (LAV), Dynamic PCA plugin, Gallery Plot plugin, and the GC × IMS Library Search software (Version 1.2.1).

3. Results and Discussion

3.1. WHC

WHC refers to the capacity of fish muscle to retain its moisture under external pressure, serving as an essential attribute that influences overall quality characteristics [26]. As shown in Figure 1A, the fresh group exhibited the highest WHC (76.83 ± 1.62%), which was significantly higher than all treated groups, representing that untreated sea bass filets maintained an intact myofibrillar structure and strong water-holding capacity of sarcoplasmic proteins. After soaking in pure water, WHC decreased to 70.71 ± 1.67%, suggesting that simple immersion may cause the leaching of proteins and soluble components, thereby reducing the amount of bound water between muscle fiber bundles and weakening water retention. pH largely determines the water-holding capacity of filets [30]. Among the three juice-marinated groups, the G group showed a WHC of 72.69 ± 1.78%, slightly higher than the C and O groups. Samples with high WHC retained more moisture throughout cooking, resulting in meat that was softer and juicier [18], whereas the L group had the lowest WHC (67.73 ± 0.60%). Nour [12] reported that this could be a result of the pH getting closer to the isoelectric point of muscle proteins, which leads to a reduction in net charge and electrostatic repulsion between proteins, subsequently decreasing the quantity of water molecules bound to the protein structure. Rupasinghe et al. [31] reported a comparable result, observing that compared with the control, meat products marinated with mango juice exhibited lower WHC.

3.2. TPA

Hardness and springiness determine the taste of fish filets [32]. The texture differences among sea bass filets under different treatments were evaluated by TPA. In terms of hardness, the fresh fish showed the highest value (580.50 ± 19.93 g), indicating that fresh filets possessed the densest gel network. After fruit juice marination, the hardness of the O and G groups was significantly lower than that of fresh filets, measuring 528.32 ± 32.88 g and 520.35 ± 23.81 g, respectively. This is in line with the results of Unal et al. [17], who improved the quality of chicken meat using grapefruit juice and attributed the effect to the acidity of the juice. Acidic conditions have been reported to promote muscle swelling and weaken protein interactions, which can contribute to softening [33]. The C group showed the lowest hardness, decreasing significantly to 426.38 ± 27.79 g, approximately 26% less than that of the F group. This suggested that soaking in pure water may cause muscle swelling and weaken interactions among myofibrils, resulting in a softer texture and reduced elasticity. The hardness of the L group fell between that of the F group and those of the O and G groups, suggesting that lemon juice marination may further increase hardness; however, considering its lower WHC, the texture may be perceived as “firmer but slightly drier”. Differences in springiness among groups were relatively small but still followed a certain pattern. The springiness of the F group was 0.62 ± 0.05, which was considerably higher than that of the C and G groups. This suggests that different juice marination treatments and pure water soaking can weaken springiness to varying degrees.

3.3. pH Value

Due to the different pH of the juice, its impact on the pH value of the fish sample is also varied. The pH in the F group fell below 7 as a result of both lactic acid production via glycolysis and its subsequent accumulation in the seafood’s early rigor mortis stage [34]. After immersing in pure water, the pH of the C group reached 7.21, which was far higher than that of the F group. This could be the result of the dilution or runoff of intracellular lactic acid caused by osmosis effects. Citrus juice marination significantly reduced pH. Fish meat marinated with lemon juice had the lowest pH (6.63 ± 0.11), followed by those marinated with grapefruit juice (6.79 ± 0.11) and orange juice (6.94 ± 0.10). This is consistent with the findings reported by Çelik et al. [35]. Augustyńska-Prejsnar et al. [36] pointed out that the decrease in pH in marinated samples was caused by the acidity of the juice and its organic acid content. The L group had the lowest pH, approaching the isoelectric point of myofibrils, which explains its lowest WHC and relatively high hardness; this may result from muscle contraction and reduced gap under acidic conditions. Considering WHC, hardness, and springiness together, grapefruit marination achieved a relatively favorable balance between “juiciness” and “firmness” of the filets.

3.4. Variation in Color

Color is an important indicator for evaluating fish quality. As illustrated in Figure 1E and Figure 2E, the L* value for the filets marinated in lemon juice was 58.30 ± 2.62, significantly higher than that of fresh filets, while ΔE reached 10.47 ± 2.86, far exceeding those of the C, O, and G groups. This indicates that lemon marination caused the most drastic color change, resulting in a visibly whiter and brighter appearance. Lemon juice’s high citric acid content may be connected to this phenomenon. Zhang et al. [37] reported that acidic conditions can induce surface protein denaturation and enhance light scattering at the muscle fiber surface, thereby increasing L*. In contrast, orange and grapefruit juice marination exerted a milder effect on L*, which is beneficial for maintaining an appearance closer to that of natural fish.

The b* value reflects the change in yellowness and blueness. Compared with fresh fish meat, the C group showed a slightly more yellow tendency; however, all three juice-treated groups showed lower b* values, indicating a shift toward blueness, with lemon-juice-marinated filets showing a significant difference from the other groups. This suggests that citrus juice marination generally reduced yellowness and imparted a “cooler-toned” light appearance, with the most pronounced effect observed for lemon juice marination.

Overall, a* values remained within a slightly negative range across treatments, but significant differences were observed in brightness and yellowness modulation. Lemon juice marination markedly increased brightness and total color difference due to high citric acid content, producing a clear “whitening” visual effect. Orange and grapefruit juice marination better maintains the color of fish, leading to a more natural light appearance that better matches consumer preferences.

3.5. Analysis of Sensory Evaluation

As shown in Figure 1F, the overall sensory scores indicated that there was no obvious difference in the comprehensive performance between group C and group F. This suggests that simple soaking provided limited improvement in odor-related sensory scores under our conditions and was associated with a lower texture-related score, while any gain in appearance may be partially attributable to the lighter color. In contrast, juice marination significantly increased the overall sensory scores: the O group (70.5 ± 1.5) and the G group (70.0 ± 0.5) were considerably higher than fresh fish meat (p < 0.05). The sensory score of the lemon-juice-marinated filet (66.5 ± 0.5) was similar to that of the F group and lower than that of the O and G groups. Combined with its color and WHC results, it can be inferred that excessive whitening and higher acidity may, to some extent, reduce consumer perception of “natural color” and “tender juiciness”, thereby limiting its sensory advantage relative to the other two groups.

3.6. Analysis of Microstructure Observation

Figure 3 depicts the microstructures of filets following marination, with the white regions denoting tissue gaps. Fresh sea bass exhibited a compact internal structure with small inter-fiber gaps, indicating intact myofibrillar structure, which favors the retention of water and soluble components. This is consistent with the higher WHC and appropriate hardness observed for the F group. After soaking in pure water, the white gaps between some muscle fibers increased noticeably, and certain fibers showed blurred edges, possibly because excessive water penetration during soaking loosened the tissue structure. This feature agrees with the significantly reduced WHC and weakened hardness and springiness in the C group, indicating that pure water soaking under our conditions may weaken the integrity of the muscle matrix. Compared with the C group, the O and G groups displayed markedly reduced inter-fiber gaps and clear fiber boundaries, with only minor fine fissures in some regions. A similar finding after marinade absorption has been reported previously, where myofibrillar expansion contributed to reduced inter-fiber spaces [38]. Considering the measured juice pH values, the lemon dilution had substantially lower pH than the orange and grapefruit. Lemon marination may destroy the protein structure due to high acidity. Jingfan et al. [39] suggested that this may cause muscle contraction and localized cracking. Studies [40] have also shown that higher acidity can induce degradation, dissolution, or collagen formation in muscle structure. In summary, a tighter muscle structure often corresponds to better texture and WHC, which is also consistent with the sensory evaluation.

3.7. SDS-Page

Figure 4 displays the SDS-PAGE profiles of different groupings. Myosin heavy chain (MHC, 220 kDa), paramyosin (PM, 100 kDa), actin (43 kDa), tropomyosin (TM, 37 kDa), and a series of myosin light chains (MLC, 15, 17, and 24 kDa) were the main protein bands seen [41]. According to Naveena et al. [42], a drop in high-molecular-weight protein bands, accompanied by a concomitant increase in low-molecular-weight protein bands, is often associated with protein breakdown in meat. Compared with other groups, the lemon juice marinated group showed decreased MHC band intensity and increased PM band intensity. This pattern may indicate structural changes in myofibrillar proteins under the lower pH conditions. However, it should be noted that band intensity can also be influenced by protein solubility as well as acid-induced aggregation or precipitation, which may alter the amount and composition of proteins recovered for electrophoresis. The actin intensity of the filets treated with lemon juice was lower than that in other treated samples, which could be related to aggregation, which prevented protein clusters from entering the gel pores, as reported previously [43]. TM and MLC showed little change after different marinades were applied. The band patterns of the O and G groups were highly similar to those of the F group, with relatively intact MHC and actin bands and only minor changes, consistent with their moderate changes in texture and WHC. It shows that orange juice and grapefruit juice do not markedly alter the major myofibrillar band patterns under our conditions while improving the texture of fish filets.

3.8. Smell Analysis

Designed to mimic the biological architecture of the human olfactory system, the electronic nose is an advanced system capable of detecting specific molecules and identifying nuanced variations in volatile compounds. This technology offers a highly efficient and rapid analytical approach that is both non-destructive and environmentally sustainable [44]. Figure 5A shows radar plots of volatile components in different groups. All five groups exhibited stronger responses in the W2W and W5S sensors, indicating that organic sulfides and nitrogen oxides were the main volatile compounds in sea bass samples [45]. As shown, significant variations in the signal intensities of sensors W1S, W1W, and W2W indicate that the differences in odor are primarily attributable to sulfides, methane, and aromatic compounds. The W1S, W1W, and W2S response values of the O, G, and L groups were higher than those in the F group, suggesting that a substantial quantity of volatile compounds was generated in the marinated fish filets. The signal intensity of most sensors in group C was lower than that in group F, showing that pure water immersion weakened the flavor of fish filets [46].

Figure 5B illustrates the PCA results derived from the electronic-nose sensor responses for sea bass filets marinated in different juices. The first and second principal components (PC1 and PC2) accounted for 58.34% and 38.58% of the variance, respectively. With a combined cumulative contribution of 96.92%, the two components successfully captured the comprehensive aroma profiles across the different sample groups [47]. On the PCA plot, group F exhibited no overlap with either the samples marinated with juice or the group soaked in pure water, suggesting that unsoaked samples possess aroma characteristics distinct from the soaked ones. In contrast, the distribution of the orange-juice-marinated group overlapped with those of groups C and L, indicating that there were no significant differences in odor profiles among these three treatments.

3.9. Taste Analysis

The electronic tongue system utilizes chemical sensors to emulate the human gustatory system, facilitating a comprehensive analysis of food taste profiles. This technology enables the quantitative assessment of six fundamental attributes: sourness, sweetness, bitterness, saltiness, umami, and astringency, offering a rapid, objective, and cost-effective alternative to sensory evaluation [48]. For E-tongue measurements, parallel runs were treated as technical replicates and were averaged within each fish to obtain a per-fish value; between-group inference was conducted using fish as the independent biological unit (n = 3 fish per group). Radar plot analysis revealed that lemon juice marination significantly attenuated saltiness while intensifying sourness, astringency, and bitterness, which was a result likely driven by its high acidity. Conversely, marination with orange and grapefruit juices enhanced sweetness to a degree while simultaneously mitigating bitterness and astringency.

As illustrated in Figure 5D, the two principal components accounted for 98.71% of the total variance. The taste variations among the five groups were primarily reflected in PC1, which contributed 90.89% to the total variance. The distinct clustering and clear separation of sample points underline the electronic tongue’s high sensitivity and effectiveness in discriminating flavor profiles in fish meat [49]. As noted by Liu et al. [50], proximity between samples on a PCA plot denotes similarity, whereas greater spatial separation signifies significant differences. Consequently, the lemon-marinated group exhibited a taste profile distinctly different from the other groups, while the overlap between the orange and grapefruit juice groups indicates a high degree of similarity in their taste characteristics. These observations emphasize that a detailed characterization of the volatile flavor profile is essential to fully clarify the divergent aromatic characteristics across the sample groups.

3.10. GC-IMS

To elucidate flavor variations in fish filets marinated under different conditions, volatile compounds of all samples were analyzed by GC-IMS. For GC-IMS, triplicate measurements were treated as technical replicates and were averaged within each fish prior to statistical analysis; fish-level values were used for inference (n = 3 fish per group). As shown in Figure 6A, using the spectrum of the F group as a reference, differential plots were generated by subtracting the spectra of other samples. If the volatile organic compounds (VOCs) are consistent between the two groups, the subtracted background appears white; blue denotes a concentration that is lower than that of the F group, while red denotes a concentration that is higher than that of the F group [26]. Figure 6 shows that peaks of high-concentration VOCs were mainly distributed at retention times of 300–1000 s, with drift times mainly at 0.6–1.35 s. Clearly, the number and intensity of red spots following citrus juice marination were greater than those of fresh fish and samples soaked in pure water, suggesting that citrus juice marination generated more VOCs.

To further analyze the effects of different juices on sea bass filets, fingerprint comparisons of VOCs were performed for each group. In Figure 6, each column shows variations in the signal intensity of a given VOC across samples, while each row reflects all selected signal peaks in a particular sample. This plot allows clear visualization of the entire volatile information in each sample and the variations among samples. A total of 47 signal peaks in all were found; due to differences in concentration and ion migration rates, some compounds appeared as both monomers (M) and dimers (D) [51]. Ultimately, 26 VOCs were identified, including six alcohols, five terpenes, five esters, four aldehydes, three ketones, one acid, one sulphocompound, and one acetal. Aldehydes such as Hexanal, 3-Methylbutanal, Propanal, and 2-Methylpropanal are frequently discussed in relation to reactions including the oxidative degradation of unsaturated fatty acids and are considered important volatile compounds in aquatic products due to their low sensory thresholds and often disagreeable odors [52]. In the present study, the aldehyde signals decreased after citrus juice marination, whereas they increased in the C group. These results suggest that citrus juice marination is associated with a reduction in aldehyde-related off-odors, while soaking in pure water may coincide with an increase in aldehyde formation under our conditions. Notably, the signal of 3-Methylbutanal was considerably enhanced in the C group and weakened in the O and L groups. This change could be related to differences in chemical reactions during marination (such as lipid oxidation and amino acid degradation). The underlying pathways responsible for these differences cannot be resolved from the present dataset. Citrus juices contain organic acids and various antioxidant components (such as phenolics) reported in the literature to potentially influence aldehyde formation via pH effects, chelating metal ions, and scavenging free radicals [53]. These mechanisms may plausibly contribute to the observed aldehyde trends; however, they were not directly assessed in this study. Some alcohols, such as 1-Butanol, 2-Methylpropanol, 1-Propanol, and Ethanol, have relatively high odor thresholds and may contribute little to product flavor. In contrast, 3-methyl-1-butanol has a comparatively low odor threshold and may contribute to filet flavor [54]. After soaking in pure water and grapefruit juice, the signal of 3-methyl-1-butanol increased. The signal of 2-Propanol, which has a pleasant odor, was lower in the C group than in the other groups. The signal of 3-Hydroxy-2-butanone was significantly enhanced after grapefruit juice treatment. According to Luo et al. [55], this compound has been discussed in relation to reactions involving polyunsaturated fatty acids and amino acids and may contribute a creamy aroma. Esters of long-chain fatty acids contribute fatty odors, whereas their short-chain counterparts impart sweet and fruity odors [56]. The signal of Ethyl butyrate was significantly enhanced in the O and G groups, and the signals of Ethyl 3-methylbutanoate and Ethyl isobutyrate were significantly enhanced in the G group, indicating that citrus juices contributed notably to enhancing sweet aroma in filets. Limonene, beta-Myrcene, and delta-3-Carene showed strong signal intensities in the O, G, and L groups, indicating their important roles in flavor formation by providing pleasant citrus aromas. In addition, gamma-Terpinene and beta-Pinene with strong signals were identified in the L group, which contribute fresh herbaceous aroma [57]. Therefore, the creation of the distinctive flavor of sea bass was facilitated by changes in VOCs during citrus juice marinating. It should be noted that oxidation markers were not measured in this study, and citrus juice compositional attributes beyond pH/°Brix (e.g., titratable acidity and polyphenols) were also not quantified. Therefore, any mechanistic interpretation linking the observed VOC shifts to oxidation inhibition or antioxidant effects remains tentative. Future work integrating oxidation indices together with broader juice compositional profiling is needed to test these hypotheses.

4. Conclusions

In this study, using E-nose, E-tongue, and GC-IMS analysis, three citrus-juice-marinated fish filets were methodically assessed for flavor profiles, sensory quality, texture, and color. The results showed that after being marinated with grapefruit juice and orange juice, the fish filets exhibited the best textural hardness while retaining the juiciness of the fish. Lemon juice showed the most pronounced whitening effect and strong flavor regulation, but was accompanied by a lower WHC under the more acidic marination conditions. GC-IMS identified 26 VOCs, showing that citrus juice marination was associated with reduced signals of several aldehydes related to odors and increased signals of fruity compounds. Grapefruit juice enhanced ester-related fruity notes (e.g., ethyl butyrate, ethyl 3-methylbutanoate, and ethyl isobutyrate), whereas lemon juice showed higher signals of citrus flavor and plant-associated terpenes (e.g., γ-terpinene and β-pinene). Overall, citrus fruit juices effectively improved the quality and flavor profile of sea bass, with orange juice and grapefruit juice being the most acceptable. These findings support the feasibility of using citrus juices as natural marinades for developing value-added sea bass products under the “no addition” formulation trend. Further work incorporating oxidation indices together with broader citrus juice compositional characterization, as well as protein-related measurements, would help verify the proposed links and clarify the underlying mechanisms.