Species‐Dependent Physicochemical Properties and Bioactive Performance of Freshwater Fish Crackers: Linking Protein–Starch Structuring and Antioxidant Mechanisms

ABSTRACT

The development of bioactive rich functional snacks from freshwater fish is increasingly accepted in society. The goal of this research was to investigate the physicochemical, antioxidant, and structural properties of three types of crackers made from three different species of fish rohu (Labeo rohita), silver carp (Hypophthalmichthys molitrix), and tilapia (Oreochromis niloticus) with the same formulation and processing methods. The crackers made from silver carp exhibited the greatest linear expansion, together with the highest phenolic content, flavonoid content, and degree of starch–protein matrix structuring. The crackers made from rohu had the greatest retention of protein and were also able to absorb the least amount of oil, indicating that they had a more compact matrix structure. Tilapia crackers demonstrated a significantly lower amount of phenolic compounds. However, their radical‐scavenging activity was greater than that of the other two fish. This suggests that heat‐induced antioxidant peptide generation may occur during the baking process. Fourier transform‐infrared (FTIR) spectroscopy confirmed protein unfolding for each of the three species, whereas x‐ray diffraction (XRD) demonstrated the presence of semicrystalline starch organization for all three species. Using principal component analysis (PC1: 71.52%; PC2: 28.48%), it was possible to show the relationship between antioxidant chemistry, color attributes, and species differentiation (independent of proximate composition). These findings provide a mechanistic framework to develop functional freshwater fish snack products based on species of fish.

1. Introduction

Over the years, there have been increasing numbers of consumers who are moving away from the traditional energy‐dense foods we have been eating for decades to items that are more nutrient‐dense and often considered to be functional snacks as they meet the needs of consumers who want to ensure they are being healthy, sustainable, and meeting their needs for convenience (Sousa et al. 2019; Boukid et al. 2022; Rahman et al. 2026). This shift has placed additional pressure on the food industry to innovate rapidly in response to evolving market trends and regulatory changes (Bresciani 2017; Kumar et al. 2021; Corte et al. 2018). In Bangladesh, freshwater fish are a major part of the diet and contribute greatly to the economy. However, the increase in fish eaten from both capture fisheries and aquaculture has not always resulted in an equal increase in intake of micronutrients, indicating a need to develop fish‐derived food products that are both value‐added and nutritionally optimized (Bogard et al. 2017; Chowdhury et al. 2026).

Fish crackers have become a widely consumed snack category, particularly in Asian markets, owing to their characteristic crisp, brittle texture and their potential as a protein‐enriched alternative to conventional starch‐based snacks (Yang et al. 2023; Ibadullah et al. 2019). Historically, fish crackers were made of starch, but they are now available as snacks fortified with protein, antioxidants, and can be used both as stand‐alone foods and as part of a complete meal (Ramesh et al. 2018). Many freshwater fish species, including rohu (Labeo rohita), silver carp (Hypophthalmichthys molitrix), and tilapia (Oreochromis niloticus), can be used to create value‐added snacks and have several advantages, including relatively low prices, year‐round availability, and favorable nutritional profiles (Shamsuzzaman et al. 2017). However, there is variation in the muscle protein composition, lipid content, and minor bioactive compounds within each of these fish species that may affect the processing behavior and final product quality.

Fried snack products are created through a combination of energy exchange (through heat and mass) during the frying stage. The interactions between gelatinization of starch, denaturation of proteins, loss of moisture from food, and the interaction or absorption of oil by protein–lipid complexes all contribute to expanding the matrix and the formation of pores, which together contribute to the final texture of the fried product (Bhuiyan and Ngadi 2024; Okur et al. 2025). Recent research shows that the structural characteristics of the food matrix determine the amount of oil absorbed during the frying process, rather than frying conditions alone, thus allowing for the possibility that factors other than temperature and time may also control oil absorption (Rani et al. 2023; Feng et al. 2024; Sivaranjani et al. 2024). Starch gelatinization is critical in producing a porous, expanded structure of the food when being fried. However, the crisp texture or final quality attribute of fried expanded snack products results predominantly from moisture loss and retrogradation of gelatinized starch, not from starch gelatinization itself (Ramesh et al. 2018; Neiva et al. 2011).

Although many researchers have studied aspects of frying (including oil absorption) and the specific mechanisms associated with frying expanded for both doughs and starches, very little is available in regard to how the properties of raw materials will influence these processes when frying a freshwater species (such as rohu, silver carp, and tilapia) with identical processing conditions. Specifically, protein or starch interactions found in muscle tissues for each species during frying are not well understood, and many times this will determine whether the muscle matrix of each species will remain fused after cooking. Although gaining a complete understanding of this individual fish species’ behavior will not give rise to new products, it will provide industry‐wide guidance on how to select and/or process raw fish materials for costs associated with producing freshwater fish cracker products, as well as develop strategies for new freshwater fish cracker product types. This study adopts a structure‐function approach to elucidate how freshwater fish species govern molecular rearrangements, semicrystalline starch organization, and antioxidant accessibility during thermal processing. By integrating Fourier transform‐infrared (FTIR) spectroscopy, x‐ray diffraction (XRD), and multivariate statistical analysis, this work links molecular‐scale interactions to macroscopic attributes, including expansion, oil absorption, and texture. The study was unique in that it included simultaneous characterization of protein–starch structuring (FTIR and XRD), antioxidant chemistry (total phenolic content [TPC], total flavonoid content [TFC], and 2,2‐diphenyl‐1‐picrylhydrazyl [DPPH]), and physicochemical quality attributes (color, expansion, oil absorption, and texture) for three commercially important species of freshwater fish using a completely standardized formulation; no published data exist for rohu, silver carp, and tilapia crackers being produced under identical processing conditions. The collection and analyses of these multiple variables within an enclosed, multivariate framework will allow for direct mechanistic comparison of the species and serve as potential guides to species‐specific product development for functional freshwater fish snacks.

2. Materials and Methods

2.1. Sourcing and Preparing Raw Materials

We collected live samples of three species of freshwater fish, tilapia, rohu, and silver carp, from a retail market near the Surma River in Sylhet, Bangladesh. Vendors stated that the fish had just been harvested from the river, and all three species were descaled, eviscerated, filleted, and washed three times in potable water (running water) to remove residual scales, slime, blood, and other debris associated with the handling of fish before they are sold on the market. The washing step was designed to reduce surface residue from fish that were harvested recently instead of removing microbes. No samples that showed visible spoilage (odor, color, and tissue firmness) were included in the study. The washed fillets were drained, weighed, and then immediately processed into fish cracker dough to avoid additional quality deterioration before being formulated into fish crackers.

2.2. Experimental Design

The fish cracker product was made from three species of freshwater fish using a standardized formulation (Table 1), which enabled comparison of the fish crackers across fish species without a formulation bias (Mao et al. 2025; Baishak et al. 2020). All ratios of ingredients within each recipe/variation were identical; therefore, any difference found between the three fish cracker products proximate composition, antioxidant activity, and structural properties could be attributed to the protein and lipid composition of the muscle from each fish species. Protein was supplied mainly from fish flesh, whereas wheat flour was the starch used for gelatinization, binding, and puffing of the fish cracker products. Minor ingredients (sugar, salt, baking powder, and monosodium glutamate [MSG]) were added at equal proportions and are provided in the recipe to ensure a consistent flavor of the three fish species, provide protein solubilization, and ensure equal puffing of the three fish crackers (Neiva et al. 2011). The umami‐enhancing substance selected for use in this investigation is MSG instead of autolyzed yeast extract. MSG provides a consistent and defined umami enhancement within all three treatment groups (defined as chemical neutrality) as well as excludes variability caused by differences in synthesis, nucleic acid content, peptide profile, or residual hydrolysis byproducts among commercially available yeast extract products (Maluly et al. 2017). Commercially available yeast extracts have advantages for enhancing umami flavor, but due to their lack of consistency in composition, they may introduce an additional variable into a controlled study in which only the fish species to be utilized will serve as the independent variable (Jinap and Hajeb 2010; Maluly et al. 2017). Moreover, at the level (0.5%, w/w) of addition to these crackers, MSG only has a positive effect on real flavor enhancement and does not affect the protein/starch matrix, antioxidant levels or other physical characteristics measured as the primary responses in this research project. The use of single compound flavor enhancers, which are chemically defined, was justified by the study's overall purpose of making a comparison on fish speciated differences in flavor. The researcher selected to use MSG due to the fact that yeast extract differs from batch to batch and as a result there would be no correlation among the different yeast extracted used from between commercial batches, whereas the MSG will remain in constant relation among treatments. The sample size used in this study was determined by the need for enough quantity of flesh (fish) to perform triplicate physicochemical, antioxidant, and structural evaluations on three different forms of flesh with as little raw material variation as possible. It is recommended that experimental formulation developed and used in the laboratory be further validated at either pilot or commercial scale to determine if they will function as intended under production scale conditions (Zanfirescu et al. 2019).

TABLE 1.

The composition of the formulation and the purpose of the ingredients used to prepare fish crackers from three freshwater species.

Ingredient	Quantity per batch (g)	Percentage of total mix (% w/w)	Functional role in cracker matrix
Fish flesh (silver carp/rohu/tilapia)	250	45.1	Primary protein source; influences texture, gel strength, and expansion during frying
Wheat flour	270	48.7	Main starch contributor; responsible for gelatinization, binding, and crisp structure
Sugar	12	2.2	Flavor enhancer; contributes to Maillard reaction and assists expansion
Salt	12	2.2	Improves protein solubility; enhances gel network formation
Baking powder	3	0.5	Chemical leavening agent; releases CO2 during frying for puff expansion
Monosodium glutamate (MSG)	3	0.5	Enhances umami flavor profile; improves consumer acceptability

2.3. Fish Crackers Preparation

Initially, fish pieces were minced, then salted and blended with half of the wheat flour to form a coarse paste. This allowed the salt to extract proteins soluble in salt from the fish and create an initial gel network. Subsequently, dissolved mixtures of sugar, baking powder, and MSG were added to the measured water and mixed in with the fish and flour mixture, followed by the addition of the remaining flour. This entire mixture was kneaded by hand for 10 min until a smooth, homogeneous, non‐sticky dough had been formed at an ambient temperature of 25–28°C. The dough was formed into cylinder‐shaped rolls approx. A length of 40 cm long × 4–6 cm in diameter and steamed at 90°C ± 2°C for 90 min. After cooking, cooling in running tap water at about 25°C occurred for 10 min, with the rolls then stored in 4°C ± 1°C for 12 h to completely set as gels. Finally, the set gels were sliced into 1–2 mm thick slices using a mechanical slicer, dried in a forced air oven at 60°C ± 2°C for 8–10 h to 10% ± 2% moisture content, and deep‐fried in soybean oil at 190°C ± 10°C for 60 ± 5 s to form the final cracker product (Baishak et al. 2020; Kamble et al. 2023). The preparation process of freshwater fish crackers is illustrated in Figure 1.

FIGURE 1.

Schematic representation of the preparation process of freshwater fish crackers. MSG, monosodium glutamate.

2.4. Physiochemical Analysis

2.4.1. Moisture Content Determination

Moisture content of the samples was examined using a moisture analyzer (Model: i‐Thermo A64M, BEL). The aluminum pans were cleaned with tissues before use in the Moisture Analysis equipment. The moisture analyzer was preheated to 105°C. A precise scale was used to acquire 5 g of a powdered version of each sample and place it into the aluminum pans. Moisture content was then determined as described by Rossa et al. (2015).

2.4.2. Determination of Water Activity

A water activity meter (Model: WA‐60A, AMTAST) was used to assess each sample's water activity at 29°C (Adeogun et al. 2025).

2.4.3. Determination of pH

Dried material of 1 g was added to 10 mL of distilled water and mixed using a homogenizer (HG‐15D & HG‐15A, DAIHAN Scientific Co. Ltd.) for 40 s. A pH of 7.0 buffer was created, and the glass electrode was calibrated at 28°C by washing it with distilled water before inserting it into the sample solution, which was used to determine the pH value of each homogenized sample using a digital pH meter (HI‐2211, Hanna Instrument). The electrode was inserted directly into the homogenized suspension for each reading, with no filtration or centrifugation step applied beforehand so that the pH value obtained reflected the bulk homogenate rather than a clarified supernatant.

2.4.4. Determination of Total Soluble Solids (TSS)

The TSS of each sample was measured using a hand refractometer, with readings recorded in °Brix. The actual TSS content was calculated by multiplying the measured values by the dilution factor. For sample preparation, 1 g of dried material was homogenized with 10 mL of distilled water for 40 s using a homogenizer (Model: HG‐15D & HG‐15A, DAIHAN Scientific Co. Ltd.). A few drops of this homogenized solution were then used for TSS determination (Zzaman et al. 2021; Sarkar et al. 2022).

2.4.5. Determination of Color

The color of each sample powder was measured using a colorimeter (Model: PCE‐CSM4, PCE Instruments). Before taking measurements, the instrument was standardized with a white ceramic plate. For each biscuit sample, three separate readings were taken from different regions to determine the a* (redness), b* (yellowness), and L* (lightness) values (León et al. 2006; Mahmud et al. 2026).

In addition to the E index, chroma (C*, representing color saturation) and hue angle (H°, representing the perceived color tone) were also calculated from the CIELAB coordinates as follows: C* = √(a*² + b*²) and H° = tan⁻¹ (b*/a*) (expressed in degrees). These indices are well‐established standard descriptors for color quality in fish and meat‐based products (Pathare et al. 2013; León et al. 2006) and are reported alongside L*, a*, and b* in Table 3 and Section 3.2. The overall color (E index) was then calculated using the following formula:

$$E=\sqrt{L^2+a^2+b^2}$$ (1)

TABLE 3.

CIELAB color values (L*, a*, and b*), chroma (C*), and hue angle (H°) of fish crackers prepared from rohu, silver carp, and tilapia (mean ± SD).

Parameter	Rohu	Silver carp	Tilapia
L* (lightness)	47.15 ± 0.30c	58.35 ± 0.25a	56.51 ± 0.28b
a* (redness)	7.40 ± 0.12c	12.33 ± 0.15a	9.63 ± 0.13b
b* (yellowness)	12.60 ± 0.20c	20.83 ± 0.18a	20.85 ± 0.22a
Chroma (C*)	14.61	24.21	22.97
Hue angle (H°)	59.6°	59.4°	65.2°

Note: Values are expressed as mean ± standard deviation (n = 3). Different superscripts (a, b, c) within a row indicate significant differences (p < 0.05) among samples according to one‐way ANOVA followed by Tukey's post hoc test.

2.4.6. Determination of Ash

Ash content of each sample was determined following the AOAC method (Thiex et al. 2012). A crucible was prepared with 3 g of cracker powder. The powder was then burned in a muffle furnace at 550°C for 6 h and allowed to cool in a desiccator to leave only grey ashes. Once cooled, the weight of the ash was recorded, and the percentage of ash was calculated using the following formula:

$$\%\mathrm{Ash}\mathrm{content}=\frac{\mathrm{Weight}\mathrm{of}\mathrm{residue}(\mathrm{g})}{\mathrm{Weight}\mathrm{of}\mathrm{sample}(\mathrm{g})}\times 100$$ (2)

2.4.7. Analysis of Crude Fat Content

Fat content of the cracker samples was determined following the method described by Lee et al. (1996). Two grams of cracker powder were placed in a separatory funnel and mixed with 25 mL of a chloroform–methanol solution (2:1 v/v). After adding 5 mL of 0.9% sodium chloride solution, the mixture was thoroughly shaken and allowed to stand for 40 min, resulting in the formation of three distinct layers. The lower chloroform layer, containing the extracted fat, was carefully collected into a pre‐weighed beaker. The extraction was repeated, and all chloroform layers were pooled. The solvent was then evaporated using a hot water bath until dryness, leaving the fat residue in the beaker. The beaker was further dried in an oven at 105°C for 20 min, cooled in a desiccator, and weighed. The percentage of fat was calculated using the following formula:

$$\%\mathrm{Fat}=\frac{\mathrm{Weight}\mathrm{of}\mathrm{residue}(\mathrm{g})}{\mathrm{Weight}\mathrm{of}\mathrm{sample}(\mathrm{g})}\times 100$$ (3)

2.4.8. Determination of Total Nitrogen and Protein

The Kjeldahl method (Aguirre 2023) was used to determine the protein content. This procedure follows AOAC Official Method 981.10, Crude Protein in Meat (block digestion method, N × 6.25) (AOAC 2023). In this method, 2 g of cracker powder, along with 25 mL of concentrated sulfuric acid and a digestion catalyst (selenium powder, potassium sulfate, and copper sulfate pentahydrate in ratios of 1:10:2), were mixed in a 250 mL Kjeldahl flask and heated at 300°C until the digested sample was clear and colorless. After obtaining clarity, 5 mL of digest was transferred to the distillation unit; 60 mL of 40% NaOH was added to the digest, and ammonia was distilled into 10 mL of 2% boric acid containing approximately four drops of a mixed pH indicator (bromocresol green and methyl red) until 60–70 mL of distillate was collected and a color change from pink to blue occurred. The distillate was then titrated with 0.01 N standardized HCl to a faint pink endpoint. The nitrogen content (as a percentage) was calculated as follows:

$$\begin{array}{ccc}\hfill \mathrm{Nitrogen}\left(\%\right)& =& \left[\left(\mathrm{bed}\mathrm{titrant}\mathrm{mL}\times \mathrm{normality}\times 14.007\times 250\right)/\right.\hfill \\ & & \left.\left(\mathrm{bed}\mathrm{weight}\times \mathrm{mL}\mathrm{digest}\times 1000\right)\right]\times 100\hfill \end{array}$$

The total protein content (as a percentage) was then determined by multiplying the nitrogen content (as a percentage) by 6.25. The Kjeldahl technique can be used to calculate the combined total nitrogen (N) from both protein and nonprotein N (N, such as from free amino acids, nucleotides, creatine, and trimethylamine oxide) in fish flesh. The relative amount of N sourced from nonprotein sources will vary depending on the species and degree of freshness of the fish. However, as all three species being compared were processed under identical conditions and from the same classification of degree of freshness, it can be anticipated that the relative amount of N sourced from nonprotein sources will be similar for all three species. Therefore, the protein values based on the Kjeldahl method can appropriately be used consistently for interspecies comparisons, rather than as absolute true protein values. This limitation will be addressed before interpreting the results of the protein content data reported in this study.

2.4.9. Determination of Carbohydrate Content

The phenol–sulfuric acid method (Masuko et al. 2005), conducted using boiling tubes, was used to measure total carbohydrates spectrophotometrically at 490 nm. This direct spectrophotometric assay was used in place of the conventional AOAC carbohydrate‐by‐difference calculation (AOAC 2023) because it quantifies total carbohydrate directly from the sample rather than as a residual of the other proximate fractions, thereby avoiding compounding of the analytical errors from the moisture, protein, fat, and ash determinations. The cracker powder (100 mg) was first immersed into a water bath and heated for 3 h; after cooling to room temperature, it was neutralized to a pH of 7 via the addition of solid sodium carbonate until foaming ceased and finally made up to 100 mL using distilled water. From that solution, 0.2 mL of sample was diluted to 1 mL using distilled water and mixed with 1 mL of 5% phenol and 5 mL of 96% sulfuric acid and incubated for 15 min at 30°C. Absorbance was measured at 490 nm to determine carbohydrate content against a glucose standard curve.

2.4.10. Quantification of Total Dietary Fiber (TDF)

The AOAC Official Method 985.29 (AOAC 2023) describes how to determine TDF's gravimetric way using acid‐alkali hydrolysis. A 500 mL flask was used, and a 10‐g cracker sample was defatted and dried and then added to the flask. The sample underwent sequential acid digestion (30 min reflux with H₂SO₄) and alkaline digestion (30 min reflux with NaOH) and was filtered between the two digestions via washed silk. The final sample residue was dried at 110°C until a constant weight (W ₁) was achieved, then incinerated for 20 min at 700°C (W ₂). The calculation of TDF is as follows:

$$\%\mathrm{Total}\mathrm{dietary}\mathrm{fiber}=\frac{W_1-W_2}{W}\times 100$$ (4)

where W is the weight of the sample.

2.5. Antioxidant Analysis of Fish Crackers Powder

The antioxidant potential content was measured through the TPC, TFC, and antioxidant activity (AC). All the samples were extracted by the method described in Saikia et al. (2015).

The samples were extracted using 80% methanol at a 1:10 sample‐to‐solvent ratio and then incubated for 90 min at 20°C in a shaking incubator (Sl‐200, Korea). The extract was centrifuged for 15 min at 3000 rpm (Model 416G, Gyrozen, Korea) following the incubation time. The supernatants were taken very carefully and used for further investigation.

2.5.1. Determination of Antioxidant Activity (DPPH)

To determine the antioxidant properties of various extracts, samples were tested for potential DPPH radical scavenging activity using the DPPH radical scavenging assay according to the method detailed in a publication by Brand‐Williams et al. (1995). Briefly, 1.4 mL of a methanolic DPPH solution (1 × 10⁻⁴ M) was mixed with 100 µL of the sample and incubated in the dark at room temperature for 30 min to allow for radical scavenging to take place.

Absorbance measurements were obtained at 517 nm using a UV–Vis spectrophotometer (Model: UV‐1800, Shimadzu Scientific Instruments, Japan). A blank solution containing 100 µL of 80% methanol and 1.4 mL of DPPH solution served as the control. Radical scavenging activity was calculated using the following equation:

$$\mathrm{Radical}\mathrm{scavenging}\mathrm{activity}(\%)=\left(\frac{A_0-A_s}{A_0}\right)\times 100$$ (5)

where $A_0$ is the absorbance of the control, and $A_s$ is the absorbance of the sample extract.

2.5.2. Determination of TPC

The TPC was measured according to the Folin–Ciocalteu colorimetric method. The procedure used is an adaptation of what Lai et al. (2009) reported with some modifications. In short, 0.5 mL of plant extract and 8.5 mL of distilled water were combined with 0.5 mL of Folin–Ciocalteu reagent and left at room temperature for 5 min. Next, after mixing, 1.0 mL of sodium carbonate (35% w/v) was added to the sample mixture, and the mixture was left at room temperature for 20 additional minutes. Absorbance was determined by a UV–Vis spectrophotometer model UV‐1800 from Shimadzu of Japan, set at a wavelength of 765 nm. Distilled water served as a blank for all measured samples, while gallic acid was the standard for the preparation of calibration curves to determine TPC values per gram of dry extract (g GAE/g dry extract).

2.5.3. Determination of TFC

The TFC of the samples was determined using the aluminum chloride colorimetric method as described by Hayet et al. (2025), with some changes. Briefly, 0.5 mL of the sample extract was combined with 1.5 mL of 95% ethanol (v/v), 2.1 mL of deionized water, and 0.1 mL of a 10% aluminum chloride solution (w/v). The reaction mixture was mixed thoroughly and then left to incubate at room temperature for 40 min before the absorbance at 415 nm was determined using a Shimadzu UV–Vis spectrophotometer UV‐1800 with deionized water as a blank. A standard curve of quercetin was used to determine the TFC in milligrams of quercetin equivalents per gram of sample (mg QE/g).

2.6. Analysis of FTIR Spectroscopy

The FTIR spectra for the fish cracker samples were collected over a range of 400–4000 cm⁻¹, using an FTIR spectrophotometer (Prestige‐21, Shimadzu Corporation, Japan) in the KBr pellet method. Each sample was scanned 32 times, with a resolution of 8 cm⁻¹ and KBr as the background, in order to identify specific functional groups in the fish cracker samples (Roy and Rhim 2020).

2.7. Linear Expansion, Oil Absorption, and Texture Hardness

The method from Kingwascharapong et al. (2024) was used to measure the fish crackers linear expansion. A Vernier caliper was used to measure the samples’ diameters. The average of three measurements taken in different directions was used to calculate the dry chip's diameter. Expanded cracker diameters were assessed using the same methodology. For every treatment, five duplicates of the measurements were made. The samples were weighed both before and after being fried in palm oil at 180–200°C, and the percentage of oil absorption was calculated following the procedure outlined by Zzaman et al. (2017). A texture analyzer (TA‐XT2i Stable Micro System) was used to measure the fish crackers’ hardness in accordance with the conditions and procedure outlined by Chudasama et al. (2019).

2.8. XRD Analysis of Starch Crystallinity

Dome et al. (2020) used a powder x‐ray diffractometer (with Cu Kα radiation; λ = 1.5406 Å) for XRD analysis on finely pulverized cracker powder. The range of 2θ is used in this process is from 5° to 40°. Relative crystallinity is based on the ratio of the integrated crystalline peak area to the total diffraction area expressed as a percentage. All test specimens were run using the same instrument parameters.

2.9. Statistical Analysis

Triplicate was used to conduct all tests with a reporting of the average and standard deviation (SD), where SD represents scatter around the average. The statistical differences between groups were determined by a one‐way analysis of variance (ANOVA) with post hoc Fisher's least significant difference (LSD) test when p ≤ 0.05 was realized between averages. All statistical calculations were performed using the 18.0 version of SPSS statistical analysis software. A principal component analysis (PCA) was conducted in order to assess the multivariate relationships of physicochemical properties, antioxidant properties, color, physical properties, and structure and to visually display species‐based discrimination. Prior to conducting PCA, all 19 variables were converted to equal scales (using z‐scores). A pairwise correlation analysis between all variables was performed, and this information was presented using a correlation matrix heatmap.

3. Results and Discussions

3.1. Proximate Analysis

Fish crackers made from L. rohita, H. molitrix, and O. niloticus had a different composition depending on the species of fish that were used in their production, with significant differences in the moisture, protein, fat, carbohydrate, and TDF content of the fish cracker samples (Table 2). It was, therefore, concluded that the choice of species of freshwater fish used to produce fish crackers will have a significant impact on the nutritional composition of the final product, as well as on nutrient retention during the frying process and the behavior of the fish cracker product during production. The proximate composition of the freshwater fish crackers samples are evident in Figure 2.

TABLE 2.

The proximate composition of deep‐fried fish crackers prepared from rohu (Labeo rohita), silver carp (Hypophthalmichthys molitrix), and tilapia (Oreochromis niloticus).

Sample	Moisture (%)	Ash (%)	Protein (%)	Fat (%)	Carbohydrate (%)	Total dietary fiber (TDF)
Rohu	5.479 ± 0.12a	3.95 ± 0.05b	15.95 ± 0.15a	30.12 ± 0.25c	42.66 ± 0.28a	1.78 ± 0.02a
Silver carp	5.167 ± 0.10b	3.88 ± 0.06c	13.60 ± 0.20b	32.08 ± 0.27b	41.27 ± 0.30b	1.45 ± 0.02c
Tilapia	4.306 ± 0.09c	3.98 ± 0.07a	13.28 ± 0.18b	33.05 ± 0.3a	40.39 ± 0.32c	1.67 ± 0.02b

Note: Values are mean ± SD (n = 3). Different superscripts within a row denote significant differences (p < 0.05) by one‐way ANOVA and Fisher's LSD test.

FIGURE 2.

Visual comparison of proximate composition among freshwater fish cracker samples. TDF, total dietary fiber. Lower case letters (a‐c) indicate significant differences among the means (p < 0.05).

3.1.1. Moisture Content and Frying Effect

A range from 4.31% to 5.48% was determined as the moisture content of the fried fish crackers that we analyzed. Moisture content was significantly lower in tilapia‐based fish crackers than in those of the other two species when compared to those containing any other fish species (p < 0.05). Moisture content at the lower end of the range demonstrates that the fish crackers produced by frying will have a fully expanded, low‐moisture content, allowing for maximum crispness and maximum shelf stability. The range of moisture contents (2%–6%) found in both the fried fish crackers and starch‐based expanded snacks are representative of the fact that all crackers were expanded from a high initial moisture content during the frying process and a rapid loss of moisture from the product due to the heat of the frying process (Maneerote et al. 2009). The difference seen between freshwater fish species with respect to moisture retention will likely be determined by the way in which the internal structure of the fish muscle holds moisture. Additionally, the loss of moisture and high degree of expansion produced from frying the fish crackers made from tilapia may suggest that the heat of the frying process produced a greater percentage of fat than the other species on a species‐specific basis, resulting in a unique moisture retention capability (Neiva et al. 2011).

3.1.2. Protein Content and Species Influence

The analysis of protein content of the fish cracker samples demonstrated a significant effect by species (p < 0.05). Rohu crackers had the highest protein content at 15.95% ± 0.15%, followed by silver carp at 13.60% ± 0.20% and tilapia at 13.28% ± 0.18%. The results presented here support previously published findings on species‐dependent protein retention trends, with rohu exhibiting the maximum protein retention as noted by Ahmed et al. (2022), because rohu has a larger initial muscle protein content and superior myofibrillar protein functionality. Neiva et al. (2011) indicated that during steaming, the proteins within the fish are bound up with the gelatinized starches to create a cohesive network (protein–starch matrix) in the crackers. Therefore, the increased amount of protein found in the rohu cracker is most likely due to the formation of a more effective protein–starch network throughout the cooking process (Liu et al. 2017).

The protein concentration in fried or cracker myofibrillar proteins appears to differ from the respective raw muscle myofibrillar protein concentrations and be dependent on initial protein concentrations in addition to factors such as extent of protein–starch network formation, protein moisture loss, oil absorption, and dilution of protein due to the co‐absorption of oil used in frying (Neiva et al. 2011; L. Cui et al. 2022). Rohu myofibrillar proteins appear to create a more thermally compact gel network compared to silver carp, and therefore, rohu myofibrillar proteins would create a gel with less pore connectivity and oil ingress during frying as compared to silver carp myofibrillar proteins (Ahmed et al. 2022; Neiva et al. 2011). Because of this difference, rohu acquires the least amount of oil during frying compared to silver carp and tilapia (2.45% ± 0.15% compared to 3.12% ± 0.22% for silver carp and 4.85% ± 0.35% for tilapia) (Table 6), and the lower amount of co‐absorbed lipid by rohu would explain the different protein concentrations in final rohu and silver carp, where rohu = 15.95% and silver carp = 13.60%, respectively, based on the described mechanism of dilution as noted by Maneerote et al. (2009) and Kaewmanee et al. (2015). The dilution caused by oil absorption helps to clarify why the proteins found in fried crackers appear to rank differently than the same fish's raw muscle protein rankings (rohu > silver carp > tilapia) (Tokarczyk et al. 2025; Neiva et al. 2011; Liu et al. 2017).

TABLE 6.

Linear expansion, oil absorption, and hardness of texture of the fish crackers.

Physical properties	Rohu	Silver carp	Tilapia
Linear expansion (%)	59.08 ± 4.19b	64.89 ± 4.53a	61.87 ± 7.57ab
Texture hardness (N/cm2)	1330.78 ± 201.76b	1789.23 ± 287.03a	1874.76 ± 178.50a
Oil absorption	2.45 ± 0.15c	3.12 ± 0.22b	4.85 ± 0.35a

Note: Values are expressed as mean ± standard deviation (n = 3). Different superscripts (a, b, c) within a row indicate significant differences (p < 0.05) among samples according to one‐way ANOVA followed by Tukey's post hoc test.

3.1.3. Lipid Content and Oil Uptake Behavior

The lipid content among the samples varied widely (p < 0.05), with the lipid contents of tilapia (33.05% ± 0.30%) and silver carp (32.08% ± 0.27%) cracker products being greater than that of the rohu cracker product (30.12% ± 0.25%). Higher values for lipid content in deep‐fried cracker products are primarily due to oil absorption during frying and not from the natural lipids contained within the raw fish (Maneerote et al. 2009; Liu et al. 2017). During frying, the rapid loss of moisture causes the formation of pores in the expanded matrix, allowing for the capillary absorption of oil into the product, particularly as the product cools down after frying. Previous research has shown lipid levels in fish‐ and rice‐based cracker products between 25% and 35%, depending on the porosity of the matrix and the frying conditions (Maneerote et al. 2009; Neiva et al. 2011). As rohu crackers have the lowest lipid level of the tested products, this could suggest that the internal structure of the cracker is denser or has less pore connectivity than the other cracker products produced under the same frying conditions, thus preventing oil from penetrating into the rohu cracker.

3.1.4. Carbohydrate Dominance and Matrix Composition

The largest proportion of macronutrients in the tested fish crackers was carbohydrate (40.39%–42.66%), primarily due to the starch added as an ingredient and serving as the main structural and expansion agent in the formulation of the fish cracker. Other studies have shown that starch provides the matrix for fish crackers and plays a role in developing and expanding during frying (Neiva et al. 2011). The higher carbohydrate content of the rohu crackers may also be a result of their reduced oil absorption and therefore an increased relative contribution from non‐lipid solids to the final fried product.

3.1.5. TDF Content

The TDF content of fish cracker products was between 1.45% and 1.78%. Rohu‐type had the highest TDF content. This is in line with the findings of other researchers that found a small amount of TDF in traditional fish cracker products made from refined starches and fish fillets unless the manufacturer added fiber‐rich materials such as plant fibers to the ingredient list (Kabir et al. 2022). Although these products may contain more protein than cereal‐based snacks, the low TDF levels indicate that there is additional room for the nutritional improvement of fish cracker products through the addition of resistant starch‐ and/or plant‐based fiber sources, which would improve the functional and healthful attributes of the products and maintain their desirable textural properties.

The proximate composition of fish crackers illustrates how protein retention, oil absorption, and nutritional quality of fish crackers are influenced by which species are selected. The fish crackers processed from rohu have more protein and lower amounts of fat compared to those processed from tilapia and silver carp, which reflects a denser protein–starch matrix that limits the amount of oil absorbed by the cracker structure (Neiva et al. 2011; Liu et al. 2017). In contrast, the crackers produced from tilapia and silver carp have a more porous structure that allows for the absorption of more oil, a phenomenon commonly associated with pore formation during frying and moisture loss (Maneerote et al. 2009). The presence of carbohydrates relative to other nutrients within all the samples reflects expansion due to starch in the matrix. Additionally, the very low amounts of dietary fiber present within the samples indicate an opportunity for enhancing the nutritional quality of the matrix. The health benefits associated with adequate amounts of dietary fiber include improved digestion and feeling full. Therefore, to improve nutrition, one could add additional ingredients high in dietary fiber into the matrix along with existing ingredients. However, care must also be taken to maintain the food's original food texture, as it has been documented that adding more fiber can result in decreased expansion and crispiness when snacks are fried and/or extruded (Sinaki and Koksel 2024). Moreover, it would appear that fat‐soluble vitamins are probably retained better within a matrix with more lipids, confirming the role of choosing species on the retention of micronutrient quality. These studies provide new mechanistic understanding regarding how species affect the matrices of fish crackers and allow for the development of practical recommendations for better nutritional and functional properties when developing fish crackers.

3.2. Color Analysis

Table 3 presents the effect of fish species on the color characteristics of fried fish crackers. Fried fish crackers made from rohu had significantly lower lightness L* (47.15, “c”), indicating that these products appeared darker than other fried fish crackers. In contrast, silver carp (58.35, “a”) and tilapia (56.51, “b”) crackers produced products with higher levels of lightness, which are more attractive to consumers. For the redness a* measurement of the fish cracker, it was highest for silver carp (12.33, “a”), intermediate for tilapia (9.63, “b”), and lowest for rohu (7.40, “c”). This redness value difference can be explained by the pigment composition of each fish species and through chemical changes that occurred during frying as a result of Maillard reactions. Similar to what was observed with the L* value, yellowness b* also followed a similar pattern, where silver carp (≈20.8, “a”) and tilapia (≈20.8, “a”) products had much greater yellowness b* than rohu (12.60, “c”), which is probably attributed to the quantity of carotenoid pigments present in the species of fish and the nonenzymatic browning that took place during frying.

Beyond the L*, a*, and b* coordinates, chroma (C*) and hue angle (H°) were also calculated from the CIELAB values because these indices more directly capture color saturation and hue tone, attributes that better reflect how consumers visually perceive the color of fried fish‐based snacks, using the calculation approach of León et al. (2006). This dual reporting of CIELAB coordinates together with chroma and hue angle is consistent with the current AMSA Meat Color Measurement Guidelines for characterizing the color of muscle‐based food products (King et al. 2023). The highest values for chroma, which measures color saturation or vividness, were found for silver carp (C* = 24.21) and tilapia (C* = 22.97). By contrast, chroma was significantly lower for rohu (C* = 14.61), indicating that the color of silver carp and tilapia crackers is far more saturated and vivid than rohu crackers. Hue angle (H° = tan⁻¹ (b*/a*)) represents the angular component of color in the CIELAB color space (0°/360° = red|90° = yellow|180° = green) and measured 59.4° for silver carp, 65.2° for tilapia, and 59.6° for rohu, placing all three sample products within the yellow/orange quadrant of the CIELAB color space. The lower level of chroma for rohu compared to silver carp and tilapia indicated that crackers produced from rohu were less saturated and darker than their silver carp and tilapia counterparts, which is consistent with their lower L* values. Therefore, chroma, hue angle, and L*, a*, and b* provide a thorough and industry‐wide accepted description of the color differences that exist between fish cracker products produced from different species.

Thus, these results are clear evidence that fish species have a significant impact on the visual quality of food products made from fish; specifically, that dried fish cracker products made from silver carp and tilapia are a more attractive choice for consumers because they have a brighter and more vibrant color than those produced from rohu. Importantly, this is the first study to evaluate the L*, a*, and b* values of rohu, silver carp, and tilapia fish cracker products that were all fried using the same frying parameters. This information will be useful in helping fish processors choose appropriate sources of fish to create visually appealing and nutritionally healthy fish‐based snack foods (Maneerote et al. 2009; Neiva et al. 2011).

3.3. pH, Water Activity (a _w) and TSS Analysis

The three different fish cracker types had different water activity levels, with tilapia being the highest at 0.43, silver carp at 0.37, and rohu at 0.34 (Table 4). Despite the fact that the present study's samples contained relatively low water activity levels when compared to intermediate moisture foods (i.e., approximately 0.6–0.85), differences among the different species based on their species show that water binding to proteins and porous matrix formation were also due to physical differences between fish proteins (i.e., where/when/how they were processed) affecting the development of porous structures during steaming and frying. Generally, the higher water activity levels of the tilapia crackers suggest a greater amount of “bound” water retained within the matrix, associated with the fact that they had a greater moisture content relative to the other species. These differences may impact sensory attributes (e.g., crispness) and lipid oxidation rate over time while in storage (Katz and Labuza 1981; Ibadullah et al. 2019). Specific to the a _w rank ordering, the values across species also mirrored the rank ordering of both oil absorption (rohu: 2.45% < silver carp: 3.12% < tilapia: 4.85%) and residual moisture content (rohu: 5.48% > silver carp: 5.17% > tilapia: 4.31%). Neither the simultaneous comparison of species nor the corresponding measure has been reported before. These observations support the concept that the density of protein–starch networks specific to species largely determines the a _w value for freshwater fish crackers. During steam cooking, rohu had a denser protein matrix (low porosity). Therefore, there are fewer capillary channels to retain residual moisture post‐frying, resulting in the lowest a _w. Conversely, tilapia crackers had the highest oil absorption and the most porous structure, so they retained more moisture than rohu and had the highest a _w values, which resulted in the greatest potential for lipid oxidation and loss of crispness during storage (Maneerote et al. 2009; Ibadullah et al. 2019; Sablani et al. 2007; Neiva et al. 2011).

TABLE 4.

Evaluation of water activity, pH, and total soluble solids of fish cracker samples following deep‐frying.

Parameter	Rohu	Silver carp	Tilapia
Water activity (a w)	0.34c	0.37b	0.43a
pH	7.03a	6.83c	6.92b
TSS (°Brix)	2a	1.8b	1.8b

Note: Values are expressed as mean ± standard deviation (n = 3). Different superscripts (a, b, c) within a row indicate significant differences (p < 0.05) among samples according to one‐way ANOVA followed by Tukey's post hoc test.

In contrast, pH values of all crackers clustered around neutral (≈6.8–7.0) and did not differ significantly among species. However, they differed significantly in pH level, with rohu having the highest pH (7.03), followed by tilapia at (6.92) and followed closely by silver carp at (6.83) (Table 4). These differences are attributed to two separate mechanisms that resulted in a species‐based pH gradient. The first is that rohu has a larger amount of protein (15.95%) than either tilapia or silver carp (Table 2), thus providing greater buffering capacity for myofibrillar protein in relation to the lower amounts of protein found in both tilapia and silver carp (Baishak et al. 2020; Neiva et al. 2011). The second is that silver carp has a relatively low pH (6.83), although it has a protein content of 13.60%, similar to that of tilapia (13.28%), this is due to the fact that silver carp contains a higher concentration of naturally occurring acidic metabolites within the muscle of the fish, which are released and concentrated during cooking, such as inosine 5′‐monophosphate (IMP) and lactic acid (Ng et al. 2024). Although there is no single pH value in this range that independently has restricting effects on the growth of microorganisms, the combination of low a _w and near‐neutral pH acts synergistically to create a hurdle effect regarding the shelf stability of these low‐moisture products (G. Cui et al. 2024; Tokarczyk et al. 2025). The relatively neutral pH range is typical for fish‐based products that lack acidic additives, and whereas pH itself does not strongly restrict microbial growth at this level, in combination with low a _w, it contributes to the hurdle effect that enhances microbial stability. In intrinsic food environments, both a _w and pH influence the growth boundaries of spoilage microorganisms, although low a _w generally exerts a stronger inhibitory effect in low‐moisture snacks (G. Cui et al. 2024; Katz and Labuza 1981).

Crackers also differed in TSS, which is measured in °Brix. Tilapia had a slightly higher TSS (2.0 °Brix) than rohu and silver carp (1.8 °Brix). The concentration of soluble carbohydrates and other small molecular compounds, such as simple sugars, water‐soluble proteins, and extractable solutes left over after frying, is largely reflected in TSS in dry snacks. The fat dilution mechanism that created the differences in protein content described in Section 3.1 is also present in this comparison, where TSS is a measure of only water‐soluble extracts that consist of water‐soluble solids, such as soluble dextrins, water‐soluble proteins, or low‐molecular‐weight flavor precursors. Because of the large amount of frying oil absorbed from the frying oil, 32.08% for silver carp and 33.05% for tilapia, these two species had a greater proportion of cracker mass than did rohu; therefore, the water‐soluble portion of solid material (i.e., TSS is diluted because the majority of the oil is not included) was less concentrated in either silver carp or tilapia than it was in rohu absolutely and proportionally (Tokarczyk et al. 2025; Kaewmanee et al. 2015). As a direct result of their lower oil absorption (2.45%) and higher non‐lipid solids contents (15.95% protein, 42.66% carbohydrate), the TSS of rohu crackers was higher (Table 2). Overall, the TSS, a _w and proximate data mutually support one another. The higher retention of soluble components linked to the moisture profile and matrix structure of tilapia crackers may be the cause of this elevated TSS, which may have an impact on the perceived sweetness and intensity of the flavor. Additionally, TSS in conjunction with a _w offers information about solute‐water interactions and matrix composition, which are crucial for overall quality attributes like flavor retention, browning reactions, and Maillard‐type changes during processing and storage (G. Cui et al. 2024; Kabir et al. 2022).

It can be seen that despite the retention of a fairly consistent pH across species, there are clear differences in the types of water activities and TSS between species. Thus, the fish species you are utilizing to manufacture your fish crackers clearly plays a strong role in determining the physical and chemical environment of the finished product, which, in turn, will affect both textural properties and potential shelf stability of your finished product (Katz and Labuza 1981; Cui et al. 2024). The aforementioned differences, when combined into one analysis using both a _w and soluble solid dynamics, provide very relevant insights when optimizing for quality and product design for fish cracker manufacturing. In summary, too much oil absorption results in lowered protein content, lowered TSS, and increased a _w; therefore, choosing the species is an important factor in determining both the physicochemical properties and the shelf life of the finished product.

3.4. Antioxidant Analysis

There were distinct species‐dependent differences in the antioxidant profiles of crackers made from three freshwater fish species: tilapia, rohu, and silver carp. TPC and TFC were highest in silver carp crackers and lowest in rohu (Table 5). Species‐specific muscle composition and protein–phenolic interactions, which are known to be impacted by thermal processing during snack production, can be blamed for these variations. Particularly in protein‐rich freshwater fish matrices, heat treatment may improve extractable phenolic‐like compounds through Maillard reaction products and protein–polyphenol conjugation (Sarmadi and Ismail 2010; Shahidi and Zhong 2015). Similar findings have been documented for freshwater fish products, where fish species and processing conditions have a significant impact on phenolic retention and transformation (Najafian and Babji 2012; Ng et al. 2024).

TABLE 5.

Total phenolic content (TPC), total flavonoid content (TFC), and 2,2‐diphenyl‐1‐picrylhydrazyl (DPPH) values of different freshwater species.

Fish species	TPC (mg GAE/g)	TFC (mg QE/g)	DPPH inhibition (%)
Tilapia	8.00 ± 0.20b	0.50 ± 0.05b	53.00 ± 1.50a
Rohu	3.00 ± 0.15c	0.30 ± 0.03b	49.00 ± 1.20b
Silver carp	9.00 ± 0.25a	2.00 ± 0.10a	35.00 ± 1.00c

Note: Values are mean ± SD (n = 3). Different superscripts within a column indicate significant differences (p < 0.05) by one‐way ANOVA and Fisher's LSD test.

Tilapia crackers showed the highest DPPH radical scavenging activity despite having lower TPC and TFC than silver carp, suggesting that phenolic quantity alone does not determine antioxidant efficacy. This disparity points to a significant role for processing‐induced bioactive peptides and particular amino acid residues, which have a potent ability to scavenge radicals. During thermal processing, freshwater fish proteins, particularly those from tilapia, are known to produce low‐molecular‐weight antioxidant peptides that may be more effective at neutralizing radicals than phenolics (Ng et al. 2024; Najafian and Babji 2012). Oxidative deterioration is one of the leading reasons for limiting the shelf life of microbiologically stable foods, and thus any observable interspecies variation in antioxidant profile, lipid levels, and matrix organization can be interpreted as an indication of their respective levels of oxidative stability (Calligaris et al. 2016). These results underscore the significance of species selection in creating functional freshwater fish‐based snack products by highlighting a crucial realization that antioxidant quality and molecular composition are more important than TPC (Ng et al. 2024). The antioxidant activity of freshwater differed according to fish species in Figure 3.

FIGURE 3.

Antioxidant activity of freshwater fish crackers derived from different species. DPPH, 2,2‐diphenyl‐1‐picrylhydrazyl; TFC, total flavonoid content; TPC, total phenolic content. Lower case letters (a‐c) indicate significant differences among the means (p < 0.05).

3.5. Linear Expansion, Oil Absorption, and Texture Hardness

The physical performance of the crackers clearly depended on the fish species used, reflecting differences in muscle composition and how fish proteins interact with wheat starch during the pre‐frying gelatinization stage. According to Table 6 silver carp crackers expanded the most (64.9%), indicating the formation of a starch–protein network capable of retaining steam efficiently during frying and promoting greater puffing. In contrast, rohu crackers expanded less but absorbed the least amount of oil, which points to the development of a comparatively compact outer crust that limits the formation of open capillaries through which oil can penetrate. Tilapia crackers showed a different behavior altogether: Although their expansion was only moderate, oil uptake was relatively high. This suggests a microstructure dominated by smaller, more rigid cells that increase mechanical resistance while simultaneously retaining frying oil within fine pores.

Oil absorption and texture variations for fish crackers result from different protein–starch matrix structural organizations. Interfacial structuring, as well as biopolymer interactions, determines the structural stability and performance of food colloidal systems, thus indirectly influencing microstructure, gel strength, and lipid reduction (Wijaya et al. 2019). Taken together, these patterns are consistent with earlier reports showing that the fish‐to‐starch ratio, extent of starch gelatinization, and the intrinsic protein and mineral content of fish muscle play a decisive role in governing puffing behavior and fat absorption in fried fish crackers. The findings also highlight an important methodological point, linear expansion alone cannot fully explain textural or oil‐uptake differences and should be interpreted alongside compositional and microstructural evidence to accurately describe the physical quality of fish‐based crackers (Ziaiifar et al. 2008; Lin et al. 2021; Nawaz et al. 2021; Ramesh et al. 2018).

3.6. FTIR Analysis

According to their FTIR spectral results, fish cracker samples made from L. rohita, H. molitrix, and O. niloticus all displayed similarities in terms of their absorption peaks corresponding to protein, lipid, and phenolic‐related compounds. Each type of fish cracker displayed substantial changes in chemical structure as a result of thermal processing. The broad peak observed in Figure 4 at approximately 3200–3600 cm⁻¹ (O–H stretching) is much stronger for silver carp and tilapia than it is for fish cracker, which demonstrates that silver carp and tilapia have more hydroxyl groups from hydrogen bonding, which contributes to their phenolics and Maillard reaction products (Sarmadi and Ismail 2010; Shahidi and Zhong 2015). The presence of peaks around 2989 cm⁻¹ (C–H stretching) can be attributed to the presence of aliphatic chains in both lipids and amino acid side groups. The narrow band observed in the 1700–1600 cm⁻¹ range (Amide I and II) indicates the presence of protein in all fish cracker samples, but tilapia exhibited stronger amide diffraction bands than silver carp and L. rohita, indicating that tilapia experienced a greater degree of protein unfolding and peptide exposure during processing (Najafian and Babji 2012).

FIGURE 4.

FTIR spectra of freshwater fish crackers prepared from different fish species. FTIR, Fourier transform infrared.

Comparison of the ∼1200–1000 cm⁻¹ region showed differences between silver carp and tilapia cracker samples with respect to their phenolic and flavonoid content and some of the strength of their C–O and C–N stretching vibrations as well, respectively, which is consistent with the greater phenolic and flavonoid content of the silver carp cracker comparative study. The greater amide and aromatic amino acid–related signal (≈879 cm⁻¹) appeared to be sharper for tilapia cracker, indicating an observable qualitative difference between both protein‐ and phenolic‐derived segments. This molecular difference supports the functionality of the antioxidant capacity of each species wherein, a greater phenolic signal does not correlate with higher free radical scavenging capacity (Najafian and Babji 2012; Ng et al. 2024). Overall, the FTIR spectra demonstrate that antioxidant capacity in fish crackers is driven more by thermally induced changes in protein and peptide conformation than by the total quantity of phenolic compounds. This integrated approach linking FTIR molecular fingerprinting with functional antioxidant performance provides a meaningful contribution to the structural characterization of functional seafood‐based snack products.

3.7. XRD Analysis

Fish cracker XRD patterns are shown in Figure 5 for produced fish crackers made from fish from the freshwater fish species, O. niloticus (tilapia), H. molitrix (silver carp), and L. rohita (rohu). The XRD patterns exhibited broad diffraction halos with weak peaks superimposed. All fish crackers analyzed had broad diffraction halos in the 2θ range of 15°–25° demonstrating a semicrystalline structure. The broad range of 2θ of the broad diffraction halos demonstrates a semicrystalline structure that is typical behavior of thermally processed protein–starch systems in which ordered crystalline regions coexist with amorphous protein matrices (Dome et al. 2020; Nawaz et al. 2021). The absence of sharp, well‐defined peaks in the XRD patterns is indicative of extensive molecular rearrangements throughout the processing, such as protein denaturation and starch gelatinization; molecular rearrangements resultant of protein denaturation and starch gelatinization will disrupt long‐term crystalline order (Ramesh et al. 2018). The differences seen in fish cracker samples’ structural organization can probably be attributed to the differences in the matrix of protein and starch created during thermal processing. Accurate geometric characterization of the food microstructure is crucial to understanding the transport of matter and changes in structure during thermal processing, as the arrangement of structural domains (on a spatial basis) is a critical factor affecting the mass and heat transfer of all materials in the matrix (Rahman et al. 2018).

FIGURE 5.

(A) X‐ray diffraction (XRD) patterns and (B) relative crystallinity of freshwater fish crackers formulated with different species.

The quantitative crystallinity analysis in Figure 5 showed that the crystallinity indices for tilapia (68.9%) and silver carp (68.8%) were comparable, whereas rohu had a lower index of 66.5%. These numerical differences are not large; however, they demonstrate a species‐specific molecular packing and interaction efficiencies between proteins, starch, and minor lipid components. The slight increase in crystalline structure found in fish crackers made from tilapia and silver carp may be due to the use of protein–protein and protein–starch interactions during thermal treatment, providing for a more efficiently connected structure (Dome et al. 2020). The lower level of crystalline structure that was found in fish crackers made from rohu suggests that they may possess a more amorphous matrix, which promotes molecular mobility and may impact functional properties (i.e., antioxidant access).

3.8. Principal Component Analysis

For the three freshwater cracker species, PCA analyzed 19 quality factors, including proximate composition, physical/chemical properties, antioxidant activity, color, physical properties, and starch crystallinity (by XRD). All data were standardized (by z‐scores) to remove potential differences due to different units of measurement before PCA was performed. The first two principal components together explained 100% of the total variability (PC1: 71.52%; PC2: 28.48%), greatly exceeding the suggested threshold of 70% cumulative variability for biological multivariate data, indicating that the PCA model accurately represents between species variation (Shahidi and Zhong 2015; Calligaris et al. 2016). Figure 6 illustrates the PCA biplot, demonstrating a clear separation of fish crackers samples according to fish species.

FIGURE 6.

Principal component analysis (PCA) biplot illustrating species‐driven differentiation of fish crackers. DPPH, 2,2‐diphenyl‐1‐picrylhydrazyl; TDF, total dietary fiber; TFC, total flavonoid content; TPC, total phenolic content; TSS, total soluble solids; XRD, x‐ray diffraction.

Antioxidant properties (TPC and TFC), color characteristics (L, a, and b), and (XRD) crystallinity/texture hardness were the major drivers for PC1, whereas proximate/physical properties were the primary contributors for PC2 (notably ash content, DPPH radical scavenging activity, oil and water absorption, water activity, and moisture). Fish used for these analyses were clustered at opposite extremes of the axes where silver carp clustered positively on PC1 (indicating a high antioxidant content) due to its comparatively greater amount of phenolic and flavonoid compounds (Shahidi and Zhong 2015) than other species; thus, it exhibited better color than did any other type of fish, which improved oxidative stability in fish products (Shahidi and Zhong 2015). Comparatively, rohu was clustered at the negative PC1 end with a moderate loading on PC2; this indicated that rohu possesses a higher protein and moisture content and a tighter protein–starch interaction, which would lead to poorer oil absorption than other species of fish (Ng et al. 2024). Comparable to silver carp's clustering, tilapia clustered positively along the positive PC2 end; this indicates that this fish exhibits high levels of DPPH scavenging activity, oil and water absorption properties, and ash content, which is commensurate with their generation of heat‐induced antioxidant peptide molecules during their respective thermal processing (Najafian and Babji 2012; Ng et al. 2024).

Fat vector orientation in opposition to the TPC and TFC vectors indicates a negative association between lipid content and antioxidant potential. Protein and moisture demonstrated predominantly orthogonal loadings to the antioxidant–color axis, confirming that proximate nutritional composition operates independently of bioactive quality. The x‐ray crystallinity and hardness texture loadings were largely positively co‐loading with silver carp's PC1 axis, indicating increased starch ordering may optimize retention of antioxidants. Therefore, the differentiation of quality of fish cracker products is based on the synergistic interplay of antioxidant chemistry, color, physicochemical structure, and starch ordering rather than simply upon proximate composition (Calligaris et al. 2016; G. Cui et al. 2024). The best use of silver carp will be as an ingredient in premium functional products; rohu will be best used for protein‐enriched products; and antioxidant supplementation in tilapia crackers will help mitigate oxidative susceptibility (Ng et al. 2024).

3.9. Pearson Correlation Heatmap

A Pearson correlation heatmap was created to assess correlations between 19 quality parameters. TPC and TFC were strongly positively correlated to each other and to all color parameters (L*, a*, and b*) and XRD crystallinity; therefore, crackers with a higher amount of bioactive compounds exhibit a higher level of color quality and a greater density of the starch–protein structural order. This is consistent with the actions of polyphenolic compounds on the modulation of the Maillard reaction and maintaining color stability during cooking (Shahidi and Zhong 2015). In tilapia crackers, the DPPH scavenging activity was positively correlated to both ash content and the ability of the cracker to absorb oil and negatively correlated to TPC and TFC; therefore, DPPH radical scavenging activity in tilapia crackers is likely due to the action of peptide‐based mechanisms and not the quantity of phenolic compounds (Najafian and Babji 2012; Ng et al. 2024).  Figure 7 illustrates the correlation matrix heatmap depicting the relationships among compositional, antioxidant, and color attributes of fish crackers.

FIGURE 7.

Correlation matrix heatmap showing relationships among compositional, antioxidant and color attributes of fish crackers. DPPH, 2,2‐diphenyl‐1‐picrylhydrazyl; TDF, total dietary fiber; TFC, total flavonoid content; TPC, total phenolic content; TSS, total soluble solids; XRD, x‐ray diffraction.

There was a moderate to strong negative correlation between fat content and TPC, TFC, and color (L* and b*) that indicated that crackers high in fat have a lower antioxidant capacity and poorer color stability, which aligns with the items being more prone to oxidative color variation than those fats have demonstrated (Calligaris et al. 2016). Both the texture (hardness) and oil absorption were positively correlated with fat. Therefore, as the amount of lipid incorporated increased, the density of the cracker also increased, and there was greater oil retention within the cracker. There was a moderate correlation between protein and moisture but only a weak correlation with antioxidant and color properties; thus, as proximate nutritional values on their own, they do not affect the bioactive quality properties as supported by the PCA results. Crackers with greater linear expansion had a positive correlation with L* and a negative correlation with oil absorption, which indicates that as crackers expand more, they become a lighter shade of color and also exhibit less oil absorption due to having a more porous microstructure (Maneerote et al. 2009). The correlation analysis supports the PCA results and supports the idea that the quality of fish cracker can be derived from the integrated antioxidant–color–structural axis and not only due to proximate nutrient values (Calligaris et al. 2016; G. Cui et al. 2024).

4. Conclusion

The central finding of this study is that fish species is the primary determinant of molecular structure, antioxidant profile, and physicochemical performance in freshwater fish crackers produced under identical formulations and processing parameters. To the best of the authors’ knowledge, this is the first study to simultaneously characterize protein–starch structuring (FTIR and XRD), antioxidant chemistry (TPC, TFC, and DPPH), and full physicochemical quality (color, expansion, oil absorption, and texture) across rohu, silver carp, and tilapia processed under controlled and identical conditions, establishing a novel species‐specific benchmark for functional freshwater fish snack development. Crackers made from silver carp had higher antioxidant retention than all other fish species, color quality greater than that of all other fish species, and greater expansion than the other species, making silver carp the most suitable for premium or functional food products (antioxidant‐enriched). Rohu fish had the highest level of protein and the lowest oil absorption, making it the most suitable for reduced‐fat proteins in snack formulations. Crackers made from tilapia had the highest radical scavenging activity (DPPH) due to the heat‐induced antioxidant peptides produced by the fish rather than phenolic compounds (reduced in antioxidant level) and exhibited the highest oil absorption, which suggests a need to supplement or implement alternative processes to improve the oxidative stability of tilapia crackers over time of storage. The results of FTIR and XRD analyses provided molecular‐level relationships between protein denaturation, phenolic–protein interactions, and starch crystallinity related to the macroscale quality of crackers. The results of multivariate analysis indicated a strong positive correlation between the antioxidant chemistry, color, and structural features of the different fish species. Overall, the results of the studies provide a species‐specific/functional basis for the informed design and development of functional snacks produced from freshwater fish.

Author Contributions

Ismat Jahan Suma: methodology, investigation, data curation, conceptualization, writing – review and editing. Abid Hassan Arnab: writing – original draft, writing – review and editing, conceptualization, data curation, formal analysis, investigation, resources, validation, visualization. Md. Hassan Bin Nabi: writing – review and editing, formal analysis, data curation, software, investigation, visualization. Wahidu Zzaman: writing – review and editing, conceptualization, supervision, project administration.

Funding

The research was funded by SUST Research Center, Shahjalal University of Science and Technology, Sylhet, Bangladesh.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.