Nutraceutical potential of Terminalia arjuna bark powder in enhancing enzymatic antioxidant defenses and disease resistance in aquaculture: implications for sustainable food safety and quality

Introduction

Aquaculture stands as a vital pillar of global food security, delivering high-quality protein and essential economic opportunities, especially in developing nations where it sustains millions of livelihoods. As the fastest-growing food-production sector, it is pivotal to achieving nutritional security and the Sustainable Development Goals (SDGs) (Meena et al., 2022a). Global fish consumption has climbed to 20.5 kg per capita, outpacing production and heightening needs for advances in feed technology and disease control (Meena et al., 2022a). Yet, the sector grapples with infectious disease outbreaks that cause 10%–15% annual production losses worldwide, intensifying economic pressures on small-scale farmers and disrupting supply chains (Meena et al., 2022c). In India, a frontrunner in inland aquaculture, Indian major carps lead freshwater output at 60%–70%, with Labeo rohita (rohu) excelling due to its fast growth, polyculture adaptability, and robust Asian market demand, positioning it as a key species for sustainable expansion (Sarkar et al., 2021). The transition from extensive to intensive culture systems has precipitated recurrent epizootics, prompting indiscriminate antibiotic use. The injudicious application of chemotherapeutics has led to the emergence of antimicrobial resistance (AMR), antibiotic residues in edible tissues, and environmental contamination via bioaccumulation, posing significant threats to human health and international trade (Meena et al., 2022b). Consequently, there is an urgent need for eco-friendly, cost-effective alternatives that bolster fish health without compromising food safety.

Traditional reliance on antibiotics and chemicals has bred antimicrobial resistance, residue contamination in fish, and ecosystem damage, prompting a pivot to eco-friendly phytogenic feed additives that boost immunity, growth, and resilience without harm (Semwal et al., 2016). The World Health Organization actively promotes plant-derived products and nutrient-rich diets to curb antibiotic and chemical residues in aquaculture produce (Dada, 2015). Herbal options stand out because they are naturally sourced, cost-effective, and pose no known hazards to fish health, human consumers, or ecosystems (Shalaby et al., 2006). These plants deliver diverse phytochemicals phenolics, glycosides, flavonoids, alkaloids, terpenoids, tannins, amino acids, minerals, saponins, steroids, and volatile oils which collectively fortify both specific and non-specific immune functions in fish (Harikrishnan et al., 2012; Awad and Awaad, 2017). Polyphenols, naturally occurring plant compounds (extractable yields typically 1%–15%), exhibit diverse biological activities; key classes including flavonoids (0.1%–5%), phenolic acids (0.5%–10%), lignans (0.01%–2%), and stilbenes enhance fish growth performance, immunity, and disease resistance, boosting aquaculture productivity and sustainability (Ahmadifar et al., 2021a). Essential oils (EOs), secondary metabolites from medicinal plants, contain bioactive components such as terpenes, terpenoids, phenylpropenes, and isothiocyanates that interact synergistically; their hydrophobic nature enables penetration of bacterial and parasitic cell membranes, disrupting structures and organelle functions (Dawood et al., 2021). In aquaculture feeds, EOs supplementation improves growth, immunity, and resistance to infections in fish and shrimp (Dawood et al., 2021). Medicinal plants and their derivatives have thus emerged as viable phytotherapeutic agents in aquaculture, characterized by high palatability, synergistic phytochemical interactions, biodegradability, and minimal residual accumulation. Bioactive constituents such as phenolics, flavonoids, tannins, and saponins scavenge reactive oxygen species (ROS), chelate transition metals, and inhibit pro-oxidant enzymes, thereby reinforcing the host antioxidant cascade (Awad and Awaad, 2017; Ahmadifar et al., 2021b). By targeting innate defenses, these bioactives elevate key enzyme activities through gene upregulation, enabling fish to withstand oxidative and pathogenic stresses (Tarnecki et al., 2019). In fish, the enhancement of enzymatic antioxidants superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPx) is a primary mechanism by which herbal immunostimulants mitigate oxidative stress and improve resistance to bacterial, fungal, and parasitic pathogens (Harikrishnan et al., 2012; Van Hai, 2015). Beyond immunity, herbal feeds stimulate appetite, enhance digestive processes via boosted digestive enzymes, and improve overall survival rates (Van Hai, 2015).

Terminalia arjuna (Roxb.) Wight and Arn., a semi-evergreen tree from India’s Combretaceae family, has enriched Ayurvedic medicine for over 3,000 years, with bark decoctions renowned for cardioprotection against angina, hypertension, heart failure, and dyslipidemia by fortifying cardiac muscle and circulation (Semwal et al., 2016; Dwivedi et al., 2019; Chaudhari and Mahajan, 2016). Packed with arjunolic acid, flavonoids like arjunolone, tannins, saponins, and minerals, it delivers antioxidant, anti-inflammatory, antimicrobial, hepatoprotective, antidiabetic, hypocholesterolemic, and gastroprotective effects, backed by clinical trials enhancing ejection fraction and endothelial function. Phytochemical investigations have identified arjunolic acid, terminic acid, ellagic acid, caffeic acid, and abundant flavonoid and tannin fractions in the bark matrix (Mandal et al., 2013). Though proven in human and animal health including diabetes, cardiotonic aid, ulcers, spermatorrhoea, leucorrhoea, cough, tumors, asthma, and inflammation its aquaculture role is nascent but promising, showing antioxidant power, antimicrobial action, better feed conversion, nutrient retention, and growth in L. rohita (Meena et al., 2021a). Recently, T. arjuna bark powder (TABP) has gained momentum as a natural additive, improving growth, feed efficiency, nutrient use, survival, and body indices in L. rohita and Heteropneustes fossilis at 0.25%–1% doses (Meena et al., 2024). It elevates immunity via lysozyme, leukocytes, and enzymes, boosting survival against Aeromonas hydrophila and Edwardsiella tarda (up to 81%), while curbing stress and oxidation; studies also note hematological gains, liver protection, and pathogen resilience, marking it as an antibiotic alternative (Meena et al., 2022a; Meena et al., 2022b; Meena et al., 2022c; Meena et al., 2024). Moreover, TABP and its solvent extracts exhibit broad-spectrum antimicrobial activity against A. hydrophila, E. tarda, Salmonella spp., Aphanomyces invadans, and the ectoparasite Argulus bengalensis (Meena et al., 2022a; Meena et al., 2022b). Notably, Mohanty et al. (2023) found T. arjuna bark extract enhanced growth, biochemistry, and immunity in H. fossilis larvae up to 0.25% inclusion, but higher 0.5% doses impaired both, highlighting safe limits amid rising scrutiny of chemical promoters in intensified feeds.

As phyto-additives rise in “green aquaculture” to cut antibiotics, costs, and ecological risks while protecting biodiversity and health (Meena et al., 2022a), data gaps persist on TABP’s optimal use for enzymatic defense during bacterial infections, particularly in L. rohita seed rearing and stinging catfish like H. fossilis. Despite these promising in vivo observations, the phytochemical rationale and in vitro antioxidant capacity of different bark solvent extracts remain critical for standardizing dosage and predicting enzymatic antioxidant responses in fish (Meena et al., 2025). The paucity of data regarding the possible use of this plant extract as a dietary modulator in fish research has instigated investigations into the seed rearing of stinging catfish. Therefore, the present study was conceptualized to explore the possible use of dietary Terminalia bark to strengthen enzymatic defense during bacterial infection, and was designed to (i) qualitatively screen the phytochemical constituents of TABP and its sequential solvent extracts, (ii) quantify total phenolic content (TPC), total flavonoid content (FLV), ferric-reducing antioxidant power (FRAP), and free-radical scavenging activity (DPPH, ABTS, and NO) across bark, leaf, and fruit matrices, and (iii) evaluate the dose-response effects of TABP supplementation (0%–1%) on antioxidant enzymes (CAT, SOD), liver integrity biomarkers, hematology, and physiological resilience in L. rohita fingerlings challenged with A. hydrophila and E. tarda, thereby identifying the optimal extractive fraction and inclusion level that correlate with high antioxidant potential and enhanced disease resistance in aquaculture.

Materials and methods

Collection and preparation of plant material

Bark, leaf, and fruit samples of T. arjuna were collected from mature trees on the ICAR-Central Inland Fisheries Research Institute campus, Barrackpore, Kolkata, India. The plant material was authenticated and voucher specimens deposited. Samples were thoroughly washed under running tap water to remove extraneous dust, air-dried overnight at 60 °C in a hot-air oven, and ground to a fine powder using a blender. The powder was sieved through a 500 µm mesh to obtain uniform particles, transferred to airtight polythene bags, and stored in a desiccator until extraction (Meena et al., 2021).

Preparation of solvent extracts

A total of 21 solvent extracts were prepared from bark (AB), leaf (AL), and fruit (AF) powders using seven solvents of increasing polarity: hexane (Hex), ethyl acetate (EtAc), chloroform (Chlo), acetone (Acet), ethanol (Etoh), methanol (Meoh), and distilled water (DW). Briefly, 100 g of dry powder from each plant part was macerated in 500 mL of solvent (1:5 w/v) in a 1 L conical flask and agitated in a shaking incubator at room temperature for 36 h. The slurry was centrifuged at 6,000 rpm for 5 min, and the supernatant was filtered sequentially through 110 mm and 90 mm Whatman No. 1 filter papers. The filtrate was concentrated under reduced pressure in a rotary vacuum evaporator at a temperature just below the boiling point of the respective solvent until approximately one-tenth of the original volume remained. The concentrated extracts were transferred to amber glass vials, capped, and stored at 4 °C in the dark until further analysis (Meena et al., 2021a; b).

Qualitative phytochemical screening

Each extract (1 g dry weight dissolved in 100 mL of its mother solvent to yield 1% w/v stock) was subjected to standard qualitative tests for the detection of phenols, flavonoids, tannins, saponins, alkaloids, and phytosteroids according to established protocols (Meena et al., 2021a; b).

In Vitro antioxidant assays

Total phenolic content (TPC)

TPC was estimated by the Folin–Ciocalteu colorimetric method. Briefly, 10 µL of extract (25–1,250 μg/mL) was mixed with 1.5 mL of Folin–Ciocalteu reagent (1:4 dilution) and 5.5 mL of triple-distilled water. After incubation in the dark, absorbance was measured spectrophotometrically, and results were expressed as µg gallic acid equivalent (GAE)/m (Meena et al., 2021a; b).

Total flavonoid content (FLV)

FLV was determined by the aluminium chloride colorimetric method and expressed as µg quercetin equivalent (QE)/mL (Meena et al., 2021a; b).

DPPH radical-scavenging activity

The stable 2,2-diphenyl-1-picrylhydrazyl (DPPH•) radical-scavenging capacity was measured at 517 nm. Inhibition percentage was calculated relative to a DPPH blank, and IC₅₀ values were derived from dose–response curves (Meena et al., 2021a; b).

ABTS radical-cation scavenging activity

The ABTS•⁺ decolorization assay was performed as described previously, and scavenging percentages were computed (Meena et al., 2021a; b).

Nitric oxide (NO) scavenging activity

NO-scavenging potential was assessed using the sodium nitroprusside–Griess reaction system, and percent inhibition was recorded (Meena et al., 2021a; b).

Ferric-reducing antioxidant power (FRAP)

The FRAP assay measured the reduction of ferric-tripyridyltriazine (Fe³⁺-TPTZ) to its ferrous form (Fe²⁺) at 593 nm. Results were expressed as µM Fe(II) equivalents per Gram of extract (Meena et al., 2021a; b).

Experimental feeds formulation

Having screened the different solevent extract we decided to use bark powder as feed supplement. To conduct this study, four experimental diets were designed to be both iso-nitrogenous (containing 30.36% crude protein) and iso-caloric (providing 423.76 kcal gross energy per 100 g), using the proximate composition values of all major feed components. The diets differed only in the level of T. arjuna bark powder (TABP) incorporated: a control diet without any TABP (0%, or 0 g/kg, labelled CT), a diet with a low inclusion of 0.1% TABP (1 g/kg, T1), a moderate level of 1% TABP (10 g/kg, T2), and the highest level tested, 1.5% TABP (15 g/kg, T3). Development and nutrient analysis of experimental diets is being depicted in Table 1.

TABLE 1. Development and nutrient analysis of experimental diets.

Data are presented as mean values based on three replicates (n = 3).

TABP, refers to Terminalia arjuna bark powder.

The vitamin-mineral premix (Premix Plus) was incorporated into the diets at the following amounts per kilogram of feed: vitamin B2 (40 mg), vitamin E (15 mg), vitamin A as retinol (33 mg), vitamin D3 as cholecalciferol (0.55 mg), vitamin K (20 mg), vitamin B6 (20 mg), vitamin B12 (0.12 µg), calcium pantothenate (50 mg), nicotinamide (200 mg), choline chloride (3,000 mg), manganese (540 mg), iodine (20 mg), iron (150 mg), zinc (100 mg), copper (40 mg), cobalt (9 mg), L-lysine (200 mg), DL-methionine (200 mg), and selenium (2.5 mg).

The gross energy values for crude protein, crude fat, and carbohydrates were estimated using caloric equivalents of 23.6, 39.5, and 17.2 kJ per Gram, respectively, as specified by Brett (1973).

** Digestible energy (in kcal per 100 g of diet) was calculated as [protein percentage × 4] + [lipid percentage × 9] + [carbohydrate percentage × 4] following the method described by Halver (1976).

All ingredients were precisely weighed and combined to achieve a homogeneous mixture. This mix was then subjected to steam cooking at a pressure of 15 psi for 15 min using a standard kitchen pressure cooker (Prestige, Stainless Steel, Kolkata, India, 1.5 L capacity). Once the cooked blend cooled to room temperature, remaining components including fish oil, the antioxidant butylated hydroxy toluene (BHT), sunflower oil, choline chloride, and a composite vitamin-mineral premix were thoroughly kneaded into the dough to ensure uniform distribution.

To form the diets into pellets, the dough was extruded through a pelletizer featuring a 1 mm die (Twin Screw Extruder Pellet Machine, KK Life Science, Chennai, India). The resulting pellets were then evenly spread on aluminum trays and dried at 60 °C in a forced-air oven (Meta-Lab Scientific Industries, Mumbai, India) for 14 h. Once dried, pellets were packed in airtight polyethylene bags and stored at 4 °C until use in feeding trials.

Phytochemical analysis of Terminalia arjuna bark reveals it is a source of several notable bioactive constituents, including catechin, ellagic acid, gallic acid (−)-epigallocatechin, and 18-β-glycyrrhetinic acid (Semwal et al., 2016; Meena et al., 2021a; b).

Growth assessment

An indoor feeding experiment was conducted over a period of 90 days. Juvenile specimens of Labeo rohita (average initial weight: 20.3 ± 0.34 g, total N = 700) were sourced from a local provider and held in a 5000-L concrete tank for a 2-week acclimation at ICAR-CIFRI, Barrackpore, Kolkata, India. For the trial, 540 individuals (average initial weight: 20.7 ± 0.34 g) were randomly distributed into four dietary groups with three replicates per group, adhering to a completely randomized design. Fish were maintained in a flow-through water system within 500-L tanks, with each group receiving the experimental diets. Feeding occurred twice a day (at around 9:30 a.m. and 5:00 p.m.), with fish receiving feed until apparent satiation. Throughout the 90-day experimental period, monthly assessments were made to monitor fish growth and check overall health. Fish displaying typical signs of disease were obtained from an aquaculture operation in East Medinipur, West Bengal, India (22.107,897°N, 87.907,583°E). Passive surveillance indicated that the site had experienced up to 40% mortality among its fish stock. Individuals presenting externally visible symptoms—such as redness on the skin and haemorrhagic lesions were carefully chosen and transported in oxygenated 50-L fiberglass tanks to the Fish Pathology Laboratory at the ICAR-Central Inland Fisheries Research Institute, Kolkata, for diagnostic evaluation. Diagnostic and post-mortem assessments were performed according to established laboratory protocols. All procedures involving animals received approval from the Institutional Animal Ethics Committee at ICAR-CIFRI, Kolkata (approval no. IAEC/2020/02). Fish displaying disease symptoms and tissue lesions were humanely euthanized using clove oil (Dabur, India) at a final concentration of 50 µL per litre of water. Intestinal tissues were collected under aseptic conditions and initially cultured in tryptone soya broth (TSB, HiMedia, Thane, India) for 24 h at 28 °C. Following incubation, serial dilutions were prepared up to 10^–6 and plated onto tryptone soya agar (TSA, HiMedia, India) for bacterial isolation. Single bacterial colonies were sub-cultured on fresh TSA plates to obtain pure isolates.

Following the 90-day feeding regimen, a bacterial challenge test was conducted over the next 10 days using Aeromonas hydrophila and Edwardsiella tarda. For this purpose, 10 fish from each tank were taken to a separate glass tank facility (N = 120). The bacterial strains were propagated in TSB at 30 °C for 24 h, then harvested and triple-rinsed with sterile phosphate-buffered saline (PBS). Cell suspensions were adjusted to a final concentration of 1.97 × 10⁸ cells/mL. Each fish received an intraperitoneal injection of 100 µL bacterial suspension, corresponding to a dose of 10⁵ cells/mL. Fish survival rates and noticeable changes in behavior and external appearance were closely observed for a period of up to 10 days post-challenge. Additional analyses were performed at the close of the experimental period.

Procedures for sampling and collection of blood and serum

To evaluate growth performance, fish were weighed following the feeding regimen. At the end of the experiment, a final set of samples was taken to measure various biological responses. Three fish were randomly chosen from each tank replicate for periodic and terminal sampling and gently sedated using clove oil at a dose of 0.015 mL per liter of water. Once anesthetized, blood was drawn from the caudal vein with a 24-gauge needle attached to a 2 mL syringe pre-treated with heparin. These blood samples were placed in plastic microcentrifuge tubes and kept at 4 °C until further examination.

For serum and hematological tests, additional fish were randomly picked from every experimental group. These individuals were anesthetized with MS-222 at a concentration of 0.1 ppm before sampling. Blood collection was performed from the caudal vein using either a 24-gauge needle or a 2 mL heparinized syringe, with samples intended for serum analysis placed directly into microcentrifuge tubes without any anticoagulant. To separate the serum, the samples were set aside to clot for 40 min on ice at an incline. After clot formation, the tubes were centrifuged for 5 min at 27,950 × g, and the resulting serum was carefully pipetted off and stored at (−20 °C) for subsequent analysis.

Assessment of serum biochemical parameters

Serum biomarkers were measured using a fully automated Transasia-Erba EM – 2000 clinical chemistry analyzer. The analysis included a comprehensive panel of parameters: serum glutamate pyruvate transaminase (SGPT, U/L), serum glutamate oxaloacetate transaminase (SGOT, U/L), amylase (AMY, U/L), lipase (LIP, U/L), albumin (ALB, U/L), total protein (PRO, U/L), globulin (Glob, U/L), albumin-to-globulin ratio (A:G), glucose (GLU, mg/dL), gamma-glutamyl transferase (GGT, U/L), mean plasma retinol (MPR, mg/dL), maximal globulin (MxGB, mg/dL), triglycerides (Trig, mg/dL), creatine kinase (CKNac, U/L), alkaline phosphatase (ALPU, U/L), bilirubin (BIT, mg/dL), lactate dehydrogenase (LDH, U/L), cholesterol (CHOL, mg/dL), and creatinine (mg/dL).

Digestive enzyme activity assays

Lipase activity

The activity of lipase (EC 3.1.1.3) was evaluated following the method described by Cherry and Crandell (1932). In this assay, the release of fatty acids over time was quantified based on the volume of sodium hydroxide (NaOH) needed to maintain a constant pH. The reaction mixture consisted of distilled water, a homogenate from the target tissue, phosphate buffer at pH 7.0, and an olive oil emulsion. This mixture was thoroughly blended and then incubated at 40 °C for 24 h. After incubation, 95% ethanol was introduced along with two drops of phenolphthalein indicator, and the solution was then titrated with 0.05 N NaOH until a stable pink color appeared. A control was prepared by adding inactivated enzyme before introducing buffer and oil emulsion. The volume of alkali (milliequivalents) used during titration reflected the lipase enzyme activity.

Amylase activity

Amylase activity was determined utilizing the Dinitrosalicylic acid (DNS) method as described by Rick and Stegbauer (1974), which measures the production of reducing sugars following the enzymatic breakdown of starch. The reaction mixture comprised 1% starch solution (w/v), phosphate buffer, and tissue homogenate. Samples were incubated at 37 °C for 30 min. DNS reagent was then added to the mixture, which was subsequently heated in a boiling water bath for 5 minutes. After the reaction mixtures cooled, they were diluted, and absorbance was measured at 540 nm. Maltose served as the standard, and amylase activity was defined by the quantity of maltose (µmoles) released per minute under these assay conditions.

Protease activity

Protease activity was measured using the method described by Drapeau (1974), which is based on the ability of the enzyme to hydrolyze casein. The assay involved incubating tissue homogenate in 0.05 M Tris-phosphate buffer (pH 7.8) containing 1% (w/v) casein at 37 °C for 5 minutes. After allowing the reaction to proceed for 10 minutes, it was terminated by adding 10% trichloroacetic acid (TCA), followed by filtration of the contents. A reagent blank was prepared by adding the homogenate to the assay mixture immediately before introducing TCA, thereby skipping the incubation step. Under these conditions (pH 7.8, 37 °C), one unit of protease activity was defined as the enzyme quantity that liberates 0.001 (A280) of acid-soluble fragments per minute.

Oxidative stress enzyme activity

Superoxide dismutase (SOD)

The activity of superoxide dismutase was assessed following the approach. This assay evaluates the enzyme’s capability to inhibit the conversion of epinephrine to adrenochrome. To conduct the assay, 50 μL of sample was combined with 1.5 mL phosphate buffer and 0.5 mL of an epinephrine solution. The change in absorbance was recorded using a Transasia Elisa reader over a 2-min period after all components were mixed thoroughly. One unit of SOD activity was defined as the amount of protein required to achieve a 50% decrease in the rate of epinephrine autooxidation.

Catalase

Catalase activity (EC 1.11.1.7) was determined by the method outlined by Takahara et al. (1960). For this assay, 50 μL of tissue homogenate was mixed into 2.45 mL of 50 mM phosphate buffer (pH 7.0), and the reaction was initiated by adding 1.0 mL of hydrogen peroxide. The decrease in absorbance at 240 nm was measured every 30 s for 2 minutes. As a blank, distilled water was used in place of hydrogen peroxide. Catalase activity was calculated as the number of micromoles of hydrogen peroxide broken down per minute per milligram of protein.

Glutathione-S-transferase (GST)

The function of glutathione-S-transferase (GST) encompasses the conjugation of glutathione to various potentially toxic substances, including a range of organic compounds such as aliphatic, aromatic, heterocyclic, epoxides, and arene oxides. The method was used to assess GST activity in the samples.

Assessment of acetylcholine esterase (AChE) activity

The activity of acetylcholine esterase (AChE, EC 3.1.1.7), an important enzyme in neurotransmission, was determined using a protocol. The assay involved combining 1.0 mL of M/15 phosphate buffer (pH 7.2), 1 mL of acetylcholine substrate solution (prepared as a 1:9 dilution, 0.004 M, pH 4.0), and 0.2 mL of tissue homogenate. This reaction mixture was incubated at 37 °C for 30 min. To terminate the enzymatic reaction, 2.0 mL of alkaline hydroxylamine solution was introduced, followed by thorough mixing. Next, 1 mL of hydrochloric acid (2:1) was added and mixed well. For the control samples, tissue homogenate was added after this step. Lastly, 1 mL of 10% ferric chloride solution was included, and the absorbance of the resulting colored solution was measured at 540 nm. During the procedure, each solution was mixed carefully at every step to ensure accuracy and to prevent the formation of air bubbles, which could affect the results.

Assessment of serum hormones

Serum cortisol

The concentration of cortisol in fish blood serum was determined using an immunoassay method, following the protocol provided by a commercially available kit (Bio-Assay Technology Laboratory, E0053Fi). Results for serum cortisol levels were reported in nanograms per milliliter (ng/mL).

Serum thyroid hormones (T3 and T4)

Levels of thyroid hormones triiodothyronine (T3) and thyroxine (T4) were measured with an enzyme immunoassay, employing kits sourced from Bio-Assay Technology Laboratory (kit codes E0054Fi and E0055Fi). Hormone concentrations were expressed as ng/mL in all serum samples.

Integrated biomarker response (IBR) approach

To assess how dietary TABP affects health-related biomarkers in L. rohita, we applied the integrated biomarker response (IBR) method, as described by Beliaeff and Burgeot (2002) and Guerlet et al. (2010). This involved computing IBR values from hematological indicators, humoral immune factors, and adaptive immunity parameters, then visualizing them on star plots where each axis represents a biomarker and the enclosed area reflects the overall response. IBR calculated as Equation 1.where

Ai = Si / 2 sin ⁡ β Si cos β + Si + 1 sin ⁡ β (1)

β = Arctan Si + 1 sin ⁡ α / Si ‐ Si + 1 cos ⁡ α and α = 2 π / n , Sn + 1 = S 1.

Here, Ai represents the area linking two consecutive scores (Si and Si+1), which are successive clockwise radius coordinates on the star plot, and n denotes the total number of biomarkers included. For each treatment, the IBR index was normalized accounting for variations in gene numbers by averaging the biomarker values. This standardization enabled direct comparisons of IBRs across treatments and parameters.

For each treatment group, biomarker values were standardized by averaging, which allowed for direct comparisons of IBR values across groups regardless of the individual biomarkers used. Star plots were used to visually represent these calculations, with each axis representing a biomarker and the area formed indicating the overall IBR for that treatment. Analysis of these plots showed that different treatments produced significantly distinct IBR patterns. This demonstrates that multi-biomarker responses in L. rohita are sensitive to varying dietary and environmental stressors. Therefore, IBR analysis is an effective tool for identifying the most responsive biomarkers and summarizing biological impacts of diet and environmental stress in fish (Beliaeff and Burgeot, 2002; Guerlet et al., 2010).

LC-MS phytochemical fingerprinting

LC-FTMS instrumentation and chromatographic conditions

Metabolomic profiling was performed using a Q-Exactive™ Plus hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific) interfaced with a heated electrospray ionization (HESI-II) source operating in both positive and negative ion modes. Chromatographic separation was achieved on a Hypersil Gold™ C18 column (2.1 mm × 100 mm, 1.9 µm particle size; Thermo Fisher Scientific). The mobile phase consisted of (A) Milli-Q water and (B) methanol, both containing 0.1% (v/v) formic acid, delivered at a flow rate of 0.3 mL min^-1. The gradient elution program was: 0–2 min, 95% A (isocratic); 2–20 min, linear gradient from 95% A to 5% A; 20–25 min, hold at 5% A; 25–26 min, return to initial conditions (95% A); 26–30 min, equilibration at 95% A; followed by a 4-min column wash at 100% A. The injection volume was 3 µL (Meena et al., 2021a).

Mass spectrometry and acquisition parameters

The ion source was operated with a capillary voltage of +4.2 kV, a capillary temperature of 340 °C, a sheath gas flow of 37 arbitrary units (N₂), an auxiliary gas flow of 13 arbitrary units (N₂), and an auxiliary gas heater temperature of 400 °C. Full-scan MS data were acquired over m/z 100–1,200 at a mass resolving power of 35,000 (FWHM at m/z 200), while data-dependent tandem MS (ddMS²) scans were triggered on the most intense ions per cycle at a resolving power of 17,500, using a dynamic exclusion window of 10 s and a microscan count of 1 (Meena et al., 2021a). The instrument was calibrated routinely using Pierce™ ESI Positive and Negative Ion Calibration Solutions (Thermo Fisher Scientific), ensuring mass accuracy within <5 ppm across the acquisition range.

Compound identification and data-processing workflow

Raw data were processed with Thermo Compound Discoverer™ software (version 2.1 SP1). The Max ID workflow was employed for unknown compound detection, elemental composition prediction, and putative annotation. Key parameters were: intensity threshold 500,000; mass tolerance 5 ppm; S/N threshold 3; and adduct ions [M + H]⁺ [M + K]⁺ [M + Na]⁺ [M + NH₄]⁺, and [M−H]^-. Putative structures were searched against the mzCloud™ spectral library (ddMS² similarity search) restricted to the compound classes Endogenous Metabolites, Natural Products/Medicine, and Steroids/Vitamins/Hormones, with a precursor mass tolerance of 10 ppm, FT fragment mass tolerance of 10 ppm, and IT fragment mass tolerance of 0.4 Da. Compound identities were assigned based on the convergence of accurate mass (<5 ppm error), isotopic pattern fidelity, and MS/MS spectral matching (Meena et al., 2021b).

Chromatographic findings in context

As described in the revised Results, base-peak chromatograms obtained in positive ESI mode revealed distinct metabolomic fingerprints for each solvent extract. The chromatographic profiles were characterized by a higher density of peaks between retention times 15–25 min, corresponding to polyphenolic and flavonoid molecular features, whereas non-polar bark extracts exhibited fewer and less intense peaks, consistent with their lower antioxidant indices. These chromatographic patterns are congruent with the quantitative phytochemical and antioxidant hierarchies reported in our complementary investigations (Meena et al., 2021a; b).

Chemometric and statistical analyses

All computations were performed using SPSS Statistics (version 26.0; IBM Corp., Armonk, NY, USA) and R (version 4.3.1; R Core Team, 2023). Statistical significance was set at α = 0.05 for the feeding-trial and challenge-study endpoints, and at α = 0.01 for the in vitro antioxidant correlations, unless stated otherwise.

Experimental unit, factors, and nesting structure. The study comprised two distinct analytical domains: (i) a cross-sectional chemometric screening of T. arjuna bark extracts, and (ii) a longitudinal feeding trial with bacterial challenge in L. rohita. For the chemometric domain, the experimental unit was the independent solvent extract (n = 21). No random factors were included because the extracts represented a fixed, observational panel of antioxidant fractions rather than a biologically replicated experimental treatment.

For the in vivo domain, a completely randomized design (CRD) was employed. The experimental unit was the replicate tank (n = 4 tanks per dietary treatment; N = 16 tanks total) to avoid pseudo-replication. Fixed factors were dietary treatment (four levels: control, CT; T1; T2; T3), sampling occasion (five levels for endocrine profiling: 30, 60, and 90 days of feeding, plus 100 days post-challenge with A. hydrophila or E. tarda; three levels for serum biochemistry and digestive enzymes: 90 days, 100-Ah, and 100-Et), and pathogen challenge type (two levels: Ah versus Et) where applicable. Random factors were tank nested within treatment (Tank/Treatment]) and, when multiple fish were subsampled per tank for non-lethal assays, individual fish nested within tank (Fish/Tank]). Tank effects were treated as random intercepts in linear mixed-effects models (LMM) to partition between-tank variance from residual error; for tank-mean analyses (pooled digestive enzyme activity), tank was included as a blocking factor nested within treatment.

Repeated-measure structure. Because the same experimental units (tanks, and in some cases individually tagged fish subjected to serial non-lethal sampling) were measured across multiple time points, data were analyzed using two-way mixed (split-plot) ANOVA models. Treatment served as the between-subjects factor, and sampling occasion served as the within-subjects (repeated-measures) factor. Significant treatment × time interactions were decomposed using simple main-effects analysis. Additionally, to align with the row-wise temporal profiles reported in Tables 2–4, one-way repeated-measures ANOVA was conducted within each treatment to compare baseline (90-day) values with post-challenge responses (100-Ah and 100-Et).

TABLE 2. Showing variations in digestive enzymes under feed trial followed by challenge study.

TABLE 3. Showing values of Enzyme of neuro-transmission under feed trial followed by challenge study.

Table five presents values of Acetylcholinesterase (AchE) activity measured under the feed trial followed by the challenge study. Two-way nested ANOVA (or linear mixed-effects model) (df = 2,12): CT (F = 4.23, p = 0.042); T1 (F = 2.89, p = 0.098); T2 (F = 3.67, p = 0.067); T3 (F = 1.45, p = 0.278). Values = mean ± SD (n = 4). Superscripts (^a,^b) = significant differences within columns (one-way ANOVA + Duncan’s test, p < 0.05). Enzyme-treatment interaction: F₃, ₃₆ = 8.12, p < 0.001.

TABLE 4. Showing variations in T3, T4 and cortisol level of treatments groups under feed trail followed by challenge study.

Post-hoc procedures. For one-way between-subject comparisons (e.g., across treatments at a single time point), Duncan’s multiple range test was used to identify homogeneous subsets, as reported in the summary tables. For pairwise comparisons involving repeated-measures factors and for simple main-effects decomposition, the Bonferroni adjustment was applied to control family-wise Type I error inflation. Tukey’s honestly significant difference (HSD) test was used as a conservative alternative when equal sample sizes and simultaneous inference across all pairwise contrasts were required.

Assumption testing and model validation. For all parametric models, residuals were examined for univariate normality using the Shapiro–Wilk test and for homogeneity of variance using Levene’s test. For repeated-measures models, sphericity was evaluated with Mauchly’s test; when violated (p < 0.05), degrees of freedom were corrected using the Greenhouse–Geisser adjustment (ε < 0.75) or the Huynh–Feldt adjustment (ε > 0.75). Linearity and homoscedasticity were verified via residual-versus-fitted plots and Q–Q plots. Outlying observations were diagnosed using standardized residuals (|z| > 3.29) and Cook’s distance (D* > 4/N); identified outliers were retained only when biologically plausible. When distributional assumptions remained violated after log-transformation, non-parametric alternatives (Kruskal–Wallis H-test with Dunn’s post hoc) were applied.

In vitro antioxidant profiling. Descriptive statistics (mean ± SD) were calculated for all six antioxidant variables across the 21 solvent extracts. Bivariate associations among DPPH, FRAP, ABTS, NO-scavenging (NO•), total flavonoids (FLV), and total phenolics (TPC) were examined using Pearson product-moment correlation coefficients (two-tailed). The magnitude of shared variance was quantified by the coefficient of determination (R ²). To assess redundancy among predictors for composite nutraceutical quality-control models, collinearity diagnostics were computed for all pairwise variable combinations. Tolerance values and variance inflation factors (VIF) were derived from the diagonal elements of the inverse correlation matrix. Multicollinearity severity was classified according to standard thresholds: VIF >10 and tolerance <0.10 (severe); VIF >5 and tolerance <0.20 (moderate); and VIF <2 with tolerance >0.50 (absent), the latter indicating acceptability for combined use in regression models.

Hierarchical multiple regression analysis was employed to evaluate the unique and incremental contributions of TPC and FLV in predicting DPPH radical-scavenging activity (the dependent variable). In Block 1, TPC was entered as the sole predictor to establish baseline explanatory power. In Block 2, FLV was added to test whether it contributed significant incremental variance beyond TPC. Standardized regression coefficients (β), unstandardized coefficients (B) with standard errors (SE), t-statistics, p-values, model R ², change in R ² (ΔR ²), and VIF for each predictor were inspected. A suppression pattern was inferred when the addition of a correlated predictor inflated the standardized coefficient of an existing predictor, produced opposite-sign beta weights, or yielded a negligible ΔR ² despite high bivariate correlation, confirming psychometric redundancy.

Integrated Biomarker Response (IBR) and curve fitting. For IBR analysis, raw biomarker values were standardized (z-scores or rank-transformed relative to the control mean) to ensure commensurable scales, then aggregated into treatment-specific composite indices and visualized as radar plots. Polynomial (second-order) regression was fitted to model dose–response trends of antioxidant enzyme activities (CAT, SOD, GST) and acetylcholinesterase across treatments, with treatment level (coded 1–4) as the predictor; model fit was assessed via R ² and the ANOVA F-test. Pearson correlations were used to quantify enzyme–growth relationships (e.g., protease versus weight gain).

Sample sizes. The antioxidant standardization panel comprised n = 21 independent bark extracts. The feeding trial and challenge study used n = 3 replicate tanks per treatment (N = 12 tanks). For tank-level analyses (digestive enzyme activity, pooled serum biochemistry), the effective sample size was n = 3 per treatment. For individual-level parameters (blood hormones, oxidative stress enzymes), n = 3 (fish/tank) were assayed at each sampling occasion, yielding 12 observations per time point across the experiment. All reported values are mean ± SD unless otherwise indicated.

Results

Phytochemical profile of T. arjuna extracts

Qualitative screening of the solvent extracts confirmed the presence of phenols, flavonoids, tannins, saponins, alkaloids, and phytosteroids across bark matrices. LC-FTMS-based nutraceutical profiling (Table 5) further resolved the bark metabolome into distinct functional classes, with polar and polar-aprotic bark extracts (ethanolic, methanolic, and acetone) yielding the richest diversity of bioactive markers. Prominent phenolic acids included gallic acid (RT 14.52 min) and ellagic acid (RT 16.45 min), while the flavan-3-ol fraction was dominated by (−)-epigallocatechin (RT 17.15 min), epicatechin (RT 17.82 min), catechin (RT 17.64 min), and epigallocatechin gallate (RT 18.26 min). The bark signature was further defined by pentacyclic triterpenoids specifically arjunolic acid (RT 23.65 min), ursolic acid (RT 24.53 min), and 18-β-glycyrrhetinic acid (RT 25.17 min) alongside flavonols such as quercetin (RT 20.41 min), quercetin-3β-D-glucoside (RT 18.33 min), and myricetin (RT 21.08 min). These findings corroborate earlier reports that T. arjuna bark is enriched with arjunolic acid, ellagic acid, and terminic acid derivatives, while establishing a comprehensive nutraceutical inventory that links specific phytoconstituents and their chromatographic coordinates to the enhancement of enzymatic antioxidant defenses (SOD, CAT, GPx) and disease resistance in aquaculture species.

TABLE 5. Nutraceutical profiling of bioactive phytoconstituents from Terminalia arjuna bark extracts.

Abbreviations: ↑ = increase/upregulation; ↓ = decrease/downregulation; Ah = Aeromonas hydrophila; Et = Edwardsiella tarda; Argulus = Argulus bengalensis; Ich = Ichthyophthirius multifiliis; antibiot. = antibiotic; antioxid. = antioxidant; anti-inflamm. = anti-inflammatory; anti-ferropt. = anti-ferroptotic; anti-pathog. = anti-pathogenic; bioavail. = bioavailable; cardioprot. = cardioprotective; chelat. = chelator; conj. = conjugated; co-ind. = co-induces; cutan. = cutaneous; cyt. = cytochrome; deplet. = depletion; develop. = development; dig. = digestive; enz. = enzymes; excitotox. = excitotoxicity; ext. = extender; ferropt. = ferroptosis; fibrot. = fibrotic; GH, growth hormone; GPx, glutathione peroxidase; GR, glutathione reductase; GST, glutathione-S-transferase; Hb = hemoglobin; hepat. = hepatic; HK, head kidney; hypox. = hypoxia; IgM = immunoglobulin M; inflamm. = inflammatory; infilt. = infiltration; Lyso = lysozyme; MDA, malondialdehyde; membr. = membrane; metab. = metabolic; metabol. = metabolic; miner. = mineralization; mitigat. = mitigation; modul. = modulator; NK, natural killer cell; Nrf2 = nuclear factor erythroid 2-related factor 2; osmoregul. = osmoregulation; oxid. = oxidative; PCO, protein carbonyl; phag. = phagocytosis; phagocyt. = phagocytosis; photoprot. = photoprotective; Prx = peroxiredoxin; qual. = quality; reproduct. = reproductive; replac. = replacement; resp. Burst = respiratory burst; resil. = resilience; ROS/RNS, reactive oxygen/nitrogen species; RBC, red blood cell; scav. = scavenger; shelf-life = shelf-life; skelet. = skeletal; stabil. = stabilization; steat. = steatosis; transamin. = transaminases.

Statistical architecture of antioxidant indices for nutraceutical standardization

Descriptive profiling of extracts revealed substantial variation across the antioxidant indices, with DPPH scavenging activity averaging 52.3% ± 14.3% inhibition, FRAP at 8.5 ± 6.9 µM Fe(II)/g, ABTS at 49.2% ± 13.7%, NO-scavenging at 48.8% ± 13.4%, total flavonoid content at 312 ± 198 µg QE/mL, and total phenolic content at 486 ± 245 µg GAE/mL. Pearson correlation analysis demonstrated that DPPH shared a strong positive association with FRAP (r = 0.693, p < 0.01), an exceptionally tight correlation with ABTS (r = 0.991, p < 0.01), and an almost identical relationship with NO-scavenging activity (r = 0.990, p < 0.01). Similarly, FRAP correlated significantly with ABTS (r = 0.681, p < 0.01) and NO-scavenging (r = 0.689, p < 0.01), while total flavonoid content showed significant positive correlations with DPPH (r = 0.774, p < 0.01), ABTS (r = 0.787, p < 0.01), NO-scavenging (r = 0.787, p < 0.01), and total phenolic content (r = 0.974, p < 0.01). Total phenolic content emerged as the most broadly correlated parameter, exhibiting significant relationships with DPPH (r = 0.830, p < 0.01), FRAP (r = 0.630, p < 0.01), ABTS (r = 0.844, p < 0.01), NO-scavenging (r = 0.845, p < 0.01), and flavonoid content (r = 0.974, p < 0.01), underscoring its central role as the biochemical driver of the bark extract antioxidant capacity.

Collinearity diagnostics further clarified the structural relationships among these variables and their suitability for quality control protocols. The total phenolic content and total flavonoid content pair exhibited severe multicollinearity (tolerance = 0.051, variance inflation factor VIF = 19.6, R ² = 94.9%), indicating that these two predictors are essentially redundant in the bark matrix. Likewise, the radical-scavenging assays showed extreme redundancy: DPPH and ABTS (VIF = 55.6), DPPH and NO-scavenging (VIF = 50.0), and ABTS and NO-scavenging (VIF >1,000, R ² = 100.0%) were all functionally interchangeable. In contrast, FRAP demonstrated acceptable independence from total phenolic content (VIF = 1.66, R ² = 39.7%) and from DPPH (VIF = 1.92, R ² = 48.0%), confirming that it captures a distinct electron-transfer mechanism complementary to the hydrogen-donation chemistry indexed by the scavenging assays.

Hierarchical multiple regression analysis, with DPPH scavenging activity as the dependent variable, corroborated these patterns and established a predictive framework for nutraceutical standardization. In Step 1, total phenolic content alone accounted for 68.9% of the variance in DPPH activity (B = 0.048 ± 0.007, β = 0.830, t = 6.86, p < 0.001, VIF = 1.00), establishing it as a robust single-proxy predictor. When total flavonoid content was entered in Step 2, the overall model explained only a marginally higher 69.2% of variance (ΔR² = 0.003), with total phenolic content showing a suppression pattern (B = 0.086 ± 0.021, β = 1.483, t = 4.10, p = 0.001, VIF = 19.55) and total flavonoid content failing to contribute significant unique variance (B = −0.024 ± 0.026, β = −0.673, t = −0.92, p = 0.370, VIF = 19.55). This suppression effect—characterized by inflated standardized coefficients and opposite signs—confirms that total flavonoid content is psychometrically redundant with total phenolic content in bark extracts and offers no incremental predictive value for radical-scavenging capacity.

Implications for nutraceutical standardization and aquaculture application

These statistical findings yield critical inferences for the deployment of Terminalia arjuna bark powder as a phytogenic feed additive (Table 6). First, total phenolic content is the principal nutraceutical determinant of enzymatic antioxidant defense capacity; its ability to explain nearly 69% of DPPH variance establishes that standardized phenolic loads in aquafeed can reliably predict TABP efficacy in sparing and up-regulating endogenous fish enzymatic antioxidants including superoxide dismutase, catalase, and glutathione peroxidase under intensive culture stress. Second, the psychometric redundancy of flavonoid content with total phenolic content (VIF = 19.6, negligible ΔR²) indicates that commercial quality control protocols can safely eliminate dual phenolic-flavonoid testing, using total phenolic content alone as the analytical marker thereby reducing formulation costs while maintaining batch consistency. Third, while the radical-scavenging assays (DPPH, ABTS, NO-scavenging) are functionally interchangeable and should be reduced to a single representative assay (preferably DPPH for methodological simplicity), FRAP indexes a complementary redox mechanism that mirrors glutathione-GPx recycling in vivo; a composite nutraceutical index combining total phenolic content and FRAP therefore provides the most robust prediction of TABP capacity to fortify the full spectrum of enzymatic antioxidant defenses. Fourth, by establishing minimum total phenolic benchmarks (e.g., ≥400 µg GAE/mL in extractive concentrates), feed manufacturers can guarantee immunomodulatory potency—up-regulating serum lysozyme, complement C3, and respiratory burst activity while suppressing cortisol-mediated immunosuppression without recourse to prophylactic antibiotics, directly addressing antimicrobial resistance and chemical residue violations in food fish. Finally, by retaining only total phenolic content and FRAP as independent quality-control indices and discarding redundant flavonoid and duplicate scavenging assays, aquaculture operations can achieve sustainable standardization of phytogenic additives, reducing antibiotic dependency, ensuring compliance with Codex Alimentarius maximum residue limits, and extending post-harvest fillet shelf-life through muscle lipid-antioxidant protection collectively advancing the One Health objectives of fish welfare, environmental stewardship, and consumer protection central to sustainable aquaculture intensification.

TABLE 6. Integrated statistical analysis of antioxidant indices from Terminalia arjuna bark powder correlations, collinearity diagnostics, and predictive regression for nutraceutical standardization in aquaculture.

**p < 0.01. Dashes (−) denote empty upper-triangle cells or unused columns. NO• = NO-scavenging activity (%); Flav = total flavonoid content (µg QE/mL); Phen = total phenolic content (µg GAE/mL); Tol = tolerance; Sev = severity; DV, dependent variable; VIF, variance inflation factor. In Word, merge Column 1 cells vertically for each panel (A, B, C) to span their respective row blocks.

Growth assessment

Serum biochemical parameters

Across the various biochemical and physiological parameters measured, the results revealed distinct patterns of response to both experimental treatments and pathogen challenges, with all p-values reported according to established statistical practice (p < 0.05 considered statistically significant).

For CKNac90, values remained comparable among groups prior to challenge (p > 0.05). However, a pronounced decrease was observed in T2 compared to the other treatments (p = 0.013). After infection with Aeromonas hydrophila (Ah) or Edwardsiella tarda (Et), CKNac90 levels increased significantly in all experimental groups (p < 0.05). Within-group analysis revealed that both CT and T3 showed significant elevations when paired values between CKNac90 and the corresponding pathogen-challenged values (CKNac100-Ah and CKNac100-Et) were compared (p < 0.05), while T2 did not exhibit such a response (p = 0.12), Figure 1. This suggests a certain resilience or suppressed sensitivity in T2 regarding pathogen-induced change. Figure 2 reflects creatinine (Creat) and GGT values. Creat90 and GGT90 both exhibited significantly lower values in T2 relative to control and T1 (p < 0.05), and a marked increase was seen in T3 (p = 0.023).

Pathogen exposure did not cause statistically significant differences among CT, T1, T2, and T3 in Creat100-Ah, Creat100-Et, GGT100-Ah, or GGT100-Et (p > 0.05), but within CT there were reliable increases in Creat90 and GGT90 post-challenge (p = 0.018).

For MgXB90, levels were similar between CT and T1 (p = 0.77), dropped sharply in T2 (p = 0.043), then increased in T3 (p = 0.029), Figure 3. Within the control, all MgXB-related parameters (MgXB90, MgXB100-Ah, MgXB100-Et) differed significantly (p < 0.05) after infection.

In T1, a significant difference was observed only between MgXB90 and MgXB100-Et (p = 0.036). MPR90 peaked in T2 with a clear statistical difference compared to the other groups (p = 0.011), while other treatments showed no significant differences (Figure 4). Notably, infection with either pathogen resulted in a consistent drop in MPR values within each group, with strong statistical evidence for these changes in CT and T1 (p < 0.05).

BIT90 remained statistically unchanged among the treatment groups (p = 0.44), but reached its highest mean in T3. Within CT and T3, a significant rise after pathogen challenge was found (BIT90 vs. BIT100-Ah, p = 0.020), Figure 5.

In T1, BIT90 differed significantly from BIT100-Et (p = 0.033), but not from BIT100-Ah.

For Trig90, values did not fluctuate significantly among CT, T1, or T2, yet a significant increase was detected in T3 (p = 0.017). CHOL90 was significantly reduced in T2 (p = 0.009), while T1 and T3 were similar. LDH90 exhibited a significant decrease in T2 (p = 0.028) and subsequent increase in T3 (p = 0.012). Within-group comparisons showed all three markers (Trig, CHOL, LDH) surged after bacterial challenge in CT (p < 0.05) (Figure 6).

For GLU90, there was no significant difference between CT and T3 or between T1 and T2 (p > 0.05), but in T3, a significant reduction was measured (p = 0.021). Both GLU100-Ah and GLU100-Et increased significantly post-infection within T3 (p = 0.019), Figure 7.

The variation in Amylase (AMY) and SCOT is represented in Figure 8. SGOT90 levels differed significantly among the majority of groups except between CT and T1 (p = 0.54). Pathogen challenge further increased SGOT in most groups compared to baseline (p < 0.05). AMY90 was lowest in T2 (p = 0.012) and highest in T3, reflecting the treatment effect, with pronounced within-group increases post-challenge (p = 0.015 in CT and T3). SGPTD90 was significantly lower in T2 compared to other treatments (p = 0.027, Figure 9), with no discernible differences among CT, T1, and T3 (p > 0.05). Both LIP90 and ALPU90 reached their highest values in T3 (p < 0.05); all three enzyme markers exhibited post-infection increases in every group, with T3 consistently showing the most prominent elevation (Figure 9).

These results collectively indicate that T2 typically exhibited lower or less responsive biomarker activities relative to other treatments, often with statistical support (p < 0.05). Conversely, T3 demonstrated recovery or heightened responses in most parameters, especially following bacterial challenge. Control and T1 groups generally showed more pronounced changes post-infection, particularly in metabolic, liver enzyme, and immune function assays. All findings are reported using exact p-values or standard significance thresholds to ensure clarity and reproducibility, in line with established guidelines for scientific reporting.

Digestive enzymes

Table 2 shows key dietary effects on enzymatic and protein biomarkers in Labeo rohita. Protease (PROT90) peaked in T2 (higher than CT/T3, similar to T1) and stayed stable post-infection. Amylase (AMY90) dropped significantly in T3 (vs. others) and persisted post-challenge, with no T1/T2/T3 differences after ET. Lipase (LIP90) rose slightly (non-significant) in T1/T2 but fell in T3 pre-infection; no major post-infection changes for either pathogen. Alkaline phosphatase (ALPU100) was highest in T2 (T2 > T1 > T3 > CT), consistent post-challenge. Control (CT) enzymes declined significantly post-infection; amylase stayed stable across groups, T1/T2 showed no protease/lipase shifts after AH (T2 also stable post-ET), and T3 affected only ALPU. In summary, T2 boosted protease/ALPU, T3 suppressed amylase, and infection reduced markers especially in CT.

Table 2 displays the average values and standard deviations for different treatments measured across various parameters. Two-way nested ANOVA (or linear mixed-effects model) ANOVA (df = 5,18): CT (F = 8.45, p < 0.001); T1 (F = 6.78, p = 0.002); T2 (F = 9.12, p < 0.001); T3 (F = 11.34, p < 0.001). Values = mean ± SD (n = 4). Superscripts (^a,^b) = significant differences within columns (one-way ANOVA + Duncan’s test, p < 0.05). Correlation: Protease vs. Growth (Pearson’s r = 0.974, R² = 0.95, n = 18, p < 0.001).

Protease correlations*: • PROT90 vs. PROT100-AH: r = 0.967, p < 0.001 • PROT90 vs. PROT100-ET: r = 0.954, p < 0.001 • PROT100-AH vs. PROT100-ET: r = 0.982, p < 0.001 Amylase correlations: • AMY90 vs. AMY100-AH: r = 0.943, p < 0.001 • AMY90 vs. AMY100-ET: r = 0.928, p < 0.001 Lipase-ALP correlations: • LIP90 vs. ALP90: r = 0.945, p < 0.001 • LIP100-AH vs. ALP100-AH: r = 0.956, p < 0.001.

Enzyme of oxidative damage

In the feeding trial, catalase activity (CAT90) showed no significant differences among the control (CT), T1, and T2 groups (p > 0.05), but was significantly elevated in T3 compared to all other treatments (p < 0.05). Following infection, CAT100-Ah activity peaked in the control group and was higher than in other groups, although there was no significant difference between CT and T3, or between T1 and T2 (Figure 10). Superoxide dismutase activity (SOD90) was found to be significantly lower in T2 compared to CT and T1 (p < 0.05), but it increased sharply in T3, becoming significantly higher than in the other groups (p < 0.05). T2 continued to show the lowest SOD activity after A. hydrophila infection when compared to the other treatments.

The response trend for Edwardsiella tarda infection was nearly identical to that observed with A. hydrophila. Additionally, glutathione S-transferase activity (GST90) reached its highest level in T3, with T2 showing a significant reduction (p < 0.05) and T3 demonstrating a marked increase compared to all other groups. Within each treatment, enzyme activities in CT, T1, and T3 were significantly higher than those in T2 for all oxidative stress markers studied. The polynomial regression equations capturing these enzyme patterns across groups were: CAT90, Y = 2.915 × ² − 12.294x + 29.285 (R ² = 0.89); SOD90, Y = 2.1357 × ² − 10.261x + 29.243 (R ² = 0.80); and GST90, Y = 0.0327 × ² − 0.1619x + 0.4042 (R ² = 0.83). These findings highlight the pronounced antioxidant enzyme responses in the T3 group and the notable suppression of antioxidant capacity in T2, both before and after infection.

Enzyme of neuro-transmission

In the feeding experiment, Acetylcholinesterase activity (AchE90) did not show any significant difference between the control (CT) and T3 or between T1 and T2 (p > 0.05). After infection with either Aeromonas hydrophila or E. tarda, the overall trend remained consistent across all groups, with no notable changes among the treatments (Table 6). However, when analyzing within-group changes, both CT and T3 experienced a significant rise in AchE activity after infection (p < 0.05), while T1 and T2 showed no significant difference (p > 0.05) following infection. These results indicate that AchE activity is particularly responsive to infection in the CT and T3 groups, while T1 and T2 remain unaffected. The relationship for AchE90 across groups was best described by the polynomial equation Y = 0.4889 × ² − 2.1271x + 4.1358, with an R ² of 0.99.

Pearson correlations

• AchE90 vs. AchE100-AH: r = 0.923, p < 0.001
• AchE90 vs. AchE100-ET: r = 0.896, p < 0.001
• AchE100-AH vs. AchE100-ET: r = 0.945, p < 0.001

T3, T4 and cortisol

Table 4 reveals hormone variations (T3, T4, cortisol) across treatments during feeding trial and challenges. Cortisol showed significant differences only in Cort30 (p < 0.05); Cort60, Cort90, Cort100-Ah, and Cort100-Et had no differences (p > 0.05), with no variation among T1/T2/T3 in Cort60/90/100-Ah. Post-infection, Cort100-Ah/Et declined up to T2 (p < 0.05) but rose in T3, signaling stress recovery. T4 differed significantly across treatments: no CT-T1 difference at T4-60; T1-T2 differed at T4-30 (p < 0.05); CT differed at T4-100-Et. In T3, T4-30/90/100-Ah/Et rose up to T2 (p < 0.05) then fell. These patterns show dynamic cortisol shifts with T3 recovery and time-specific T4 responses post-infection.

Table 4 summarizes the mean ± standard deviation values for triiodothyronine (T3), thyroxine (T4), and cortisol levels recorded across different treatment groups during the feed trial and subsequent challenge phase.

Two-way nested ANOVA (or linear mixed-effects model ANOVA (df = 4,16): Cortisol: CT (F = 18.45, p < 0.001); T1 (F = 6.78, p = 0.003); T2 (F = 25.67, p < 0.001); T3 (F = 3.12, p = 0.045) T3: CT (F = 12.34, p < 0.001); T1 (F = 4.56, p = 0.021); T2 (F = 8.23, p < 0.001); T3 (F = 7.89, p < 0.001). T4: CT (F = 9.67, p < 0.001); T1 (F = 5.34, p = 0.012); T2 (F = 11.45, p < 0.001); T3 (F = 6.78, p = 0.002). Values = mean ± SD (n = 4). Superscripts (^a,^b,^c,^d) = significant differences within columns (one-way ANOVA + Duncan’s test, p < 0.05). Hormone-treatment interaction: F₃, ₄₈ = 14.23, p < 0.001.Strongest Correlations: • cort (Ah) vs. cort (Et): r = 0.945, R ² = 0.893, p < 0.001 • T4_60 vs. T4_90: r = 0.923, R ² = 0.852, p < 0.001 • T3_60 vs. T3_90: r = 0.823, R ² = 0.677, p < 0.001 Stress-Growth Relationships: • cort60 vs. T3_60: r = −0.712, R ² = 0.507, p < 0.01 • cort90 vs. T4_90: r = −0.654, R ² = 0.428, p < 0.01.

The integrated biomarker response (IBR) analysis

IBR serum

The Integrated Biomarker Response (IBR) analysis was used to assess serum biochemical parameters in Labeo rohita following different treatments (CT, T1, T2, and T3). Under both Aeromonas hydrophila (AH) and Edwardsiella tarda (ET) challenges, radar plots were produced for each parameter to visualize biomarker changes. Out of 30 biomarkers analysed including ALP, AMY, BIT, CHO, CKN, CRT, GGT, GLU, LDH, LIP, MgX, SGP, SGT, and Trig showed clear increases. AMY-AH, AMY-ET, BIT-AH, BIT-ET, GLU-AH, GLU-ET, and LDH-ET had larger, filled radar regions, especially after T2 and T3 treatments.

Figure 11 depicts the Integrated Biomarker Response (IBR) analysis of three serum biochemical parameters in L. rohita during a 90-day feeding trial across four groups—control (CT), T1, T2, and T3 followed by A. hydrophila (AH) and E. tarda (ET) infection challenges. Radar plots, labeled with “90″for post-feeding trial data, display standardized biomarker scores arranged symmetrically to form treatment-specific polygons, where larger areas signal greater physiological modulation and smaller ones indicate serum stability.

Enzymes like BIT90, GGT90, and SGOT90 showed prominent activation, especially in T2 and T3, reflecting diet-induced liver function and metabolic shifts. Moderate responses appeared in ALPU90, AMY90, CRT90, GLU90, and LIP90, while CKNAC90, SGPTD90, and TRIG90 remained stable, suggesting effective homeostatic regulation. Thus, these visualizations underscore dietary supplementation’s role in driving systemic biochemical responses during feeding and post-challenge phases.

Figure 12 Presents the IBR values of serum biomarkers under the challenge experiment, with radar plots depicting differential response intensity among treatments using four standardized temporal axes (T0–T3) and radial scaling from 0 to 4. Groups exposed to Aeromonas hydrophila (AH) and Edwardsiella tarda (ET) showed marked expansion or contraction along specific biomarker axes, where polygon areas reflect cumulative stress response magnitude and demonstrate relative protective effects of treatments during the challenge phase. After bacterial exposure, enzymes AMY, BIT, GGT, LDH, and SGP rose markedly, especially in T2/T3, signaling strong immunometabolic activation with large, filled radar regions, while GGT, LDH, and BIT consistently responded as key indicators of nutrition-immunity interplay. CHO, CRT, and MgX showed moderate changes with intermediate radar shifts in ALP-AH, CKN-ET, CRT-ET, LIP-ET, CKN-AH, GLU-AH/ET, and MgX-AH tied to challenge type, whereas stable or small plots characterized CHO-AH/ET, GGT-AH, Trig-AH, SGP, SGT, LIP, and GGT-ET across groups. Similar AH/ET patterns for BIT, SGP, and LDH suggest comparable reactions to distinct pathogens, with T2/T3 driving the strongest, most varied changes through expanded plots that enhance metabolic performance and disease resistance in Labeo rohita, while some biomarkers changed little regardless of treatment or infection status.

IBR of cortisol, T3 and T4

Figure 13 presents the Integrated Biomarker Response (IBR) analysis of endocrine stress-related biomarkers cortisol, triiodothyronine (T3), and thyroxine (T4) in Labeo rohita across treatment groups (CT, T1, T2, T3) under 90-day baseline, Aeromonas hydrophila (AH), and Edwardsiella tarda (ET) challenges. The radar plots reveal integrated endocrine response patterns using four standardized temporal axes with radial scaling from 0 to 4, where expansion along the cortisol axis indicates stress activation and modulation of thyroid hormones reflects metabolic adjustments. Cortisol showed marked polygon expansion under AH/ET challenges, indicating elevated stress responses especially pronounced in T2/T3 groups, while remaining stable with compact plots at baseline conditions. T3 displayed minimal changes across all conditions, with consistently compact radar plots indicating tight homeostatic regulation unaffected by dietary treatments or bacterial infection. In contrast, T4 exhibited dynamic elevation with broad radar plots, strongest in T2/T3 groups at both baseline and challenge conditions, linked to boosted metabolism and enhanced immune competence. The calculated IBR values summarize these combined endocrine alterations, allowing clearer interpretation of treatment effects on hormonal balance and stress physiology. Thus, cortisol and T4 proved sensitive to dietary treatments and pathogenic stress, while T3 remained stable; T2/T3 consistently enhanced these endocrine and stress pathways, demonstrating optimized hormonal conditioning for improved stress resilience in L. rohita.

IBR for digestive enzymes

Figure 14 illustrates the Integrated Biomarker Response (IBR) calculated from digestive enzyme activities amylase (AMY), lipase (LIP), and protease (PROT) in L. rohita across treatment groups (CT, T1, T2, T3) after 90-day feeding, A. hydrophila (AH), and E. tarda (ET) challenges. The radar plots utilize four standardized temporal axes with each axis representing a specific digestive enzyme, where polygon area reflects overall digestive functional status and calculated IBR values provide an integrated assessment of digestive capacity rather than relying on individual enzyme interpretation alone. Amylase showed stable activity with small, compact radar plots unaffected by dietary treatments or bacterial infection, indicating consistent carbohydrate digestion across all conditions. Lipase increased notably under AH/ET challenges, displaying larger polygon areas especially in T2/T3 groups, with moderate expansion observed post-90 days feeding, reflecting enhanced lipid utilization during stress and improved nutritional conditioning. Protease displayed the strongest and widest radar responses across all conditions, proving highly sensitive to both dietary modulation and pathogenic stress. Treatments that enhance digestive efficiency particularly T2 and T3 are represented by balanced and moderately expanded polygons for lipase and protease, whereas reduced or stable enzyme activities result in contracted or unchanged shapes as seen with amylase. These patterns highlight that T2/T3 treatments significantly enhance lipase and protease activities under both feeding and bacterial challenge conditions, supporting better nutrient digestion, protein utilization, and adaptive stress responses, while amylase maintains physiological stability regardless of experimental manipulation.

IBR of ache

The Integrated Biomarker Response (IBR) analysis revealed distinct patterns across three Ache treatment endpoints (Figure 15). ACH-90 showed a significant selective response, characterized by strong deviation along a single biomarker axis, indicating a targeted effect of Ache on this specific parameter under baseline conditions. Conversely, ACH-AH displayed minimal variation with a compact, centralized radar plot, suggesting this endpoint remained largely unaffected by the treatment. ACH-ET exhibited the most pronounced response, with broad multi-axial activation patterns indicating systemic and multifactorial impacts of Ache under challenge conditions. These findings demonstrate that Ache produces differential biomarker responses: ACH-90 and ACH-ET are highly responsive, while ACH-AH maintains physiological stability.

IBR of oxidative enzymes

Radar plots (Figure 16) display the Integrated Biomarker Response (IBR) for three oxidative stress enzymes catalase (CAT), superoxide dismutase (SOD), and glutathione S-transferase (GST)across treatment groups (CT, T1, T2, T3) under 90-day feeding, Aeromonas hydrophila (AH) infection, and Edwardsiella tarda (ET) infection, with each enzyme represented by three plots showing infection-specific and baseline responses. CAT exhibited the strongest response with large radial expansions in CAT-AH and CAT-ET, particularly in T2 and T3, and elevated constitutive activity in CAT-90, indicating high sensitivity to both nutrition and infection. SOD showed moderate responsiveness with subtle increases in SOD-AH and SOD-ET limited to T2 and T3, while SOD-90 remained modest and uniform across groups. GST demonstrated minimal changes with nearly flat plots across all conditions, indicating stable activity regardless of treatment or infection. These patterns highlight CAT as the primary antioxidant responder, SOD as moderately inducible, and GST as constitutively stable, reflecting their distinct roles in adaptive versus baseline antioxidant defense in Labeo rohita.

Discussion

Serum enzymes

Serum biochemical and enzymatic parameters provided valuable information on how dietary Terminalia arjuna bark powder (TABP) influenced Labeo rohita exposed to bacterial challenge. In the group receiving the moderate dose (T2), concentrations of creatine kinase (CK-Nac), creatinine, γ-glutamyl transferase (GGT), lactate dehydrogenase (LDH), amylase (AMY), lipase (LIP), alkaline phosphatase (ALP), bilirubin (BIT), triglycerides (Trig), and cholesterol (CHOL) were generally lower than those in the control. This pattern indicates that TABP at this level contributed to stabilizing liver, kidney, and metabolic functions, in agreement with earlier reports that T. arjuna extracts enhance antioxidant defenses and offer hepatoprotection in both fish and mammalian models (Manna et al., 2007; Mandal et al., 2013; Irshad et al., 2023; Meena et al., 2024). The immunomodulatory action of TABP has also been associated with upregulation of key immune-related genes such as Mx, ISG15, and STAT1, with peak expression observed at moderate supplementation levels around 10–11.5 g/kg feed (Mohanty et al., 2023; Ahmadifar et al., 2021b).

In contrast, fish fed the higher inclusion level (T3) exhibited a marked post-challenge rise in nearly all measured serum enzymes, including SGOT (AST), SGPT (ALT), GGT, LDH, AMY, LIP, and ALP. Elevated serum enzyme activity is generally interpreted as a negative outcome because it usually reflects leakage from damaged tissues and necrotic cells rather than a beneficial activation of enzyme systems (Harikrishnan et al., 2011; Harikrishnan and Heo, 2025). These results suggest that while a moderate dose of TABP supported metabolic resilience, excessive supplementation placed functional strain on vital organs and led to hepatocellular disruption. Similar dose-dependent toxicity of plant-based extracts has been reported by Mlozi et al. (2020), who showed that Tephrosia vogelii induced liver cell necrosis in rats at 2000 mg/kg inclusion, and by Mohanty et al. (2023) for T. arjuna bark extract in stinging catfish (Heteropneustes fossilis), where inclusion levels above 0.25% in the diet began to impair certain physiological parameters despite overall immunomodulatory benefits.

The sharp increase in alanine and aspartate transaminases at T3 is particularly noteworthy, as ALT and AST are well-established indicators of tissue damage in fish, especially in the liver, kidney, and gills (Roychowdhury et al., 2024). Elevated activities of these enzymes are associated with enhanced gluconeogenesis, mobilization of amino acids, and diversion of keto acids toward the tricarboxylic acid (TCA) cycle during stress, mechanisms described by Chatterjee et al. (2006). Analogous responses have been observed under crowding, thermal, or handling stress in carp (Tejpal et al., 2009; Pakhira et al., 2021). ALP followed a similar trend in the T3 group; although this enzyme plays important roles in brush-border digestion, membrane transport, and protein synthesis (Nunes, 2024), abnormally high serum activity is regarded as pathological. Pradhan (2012) likewise reported elevated ALP in L. rohita fed Chlorella after challenge with A. hydrophila, reinforcing the view that infection combined with dietary additives can disturb enzymatic homeostasis (Meena et al., 2024; Pradhan et al., 2023).

LDH, another sensitive marker of tissue injury (Ramesh et al., 1993; Harikrishnan and Heo, 2025), increased sharply in the T3 fish. This rise often coincides with the shunting of pyruvate toward anaerobic glycolysis, a characteristic feature of metabolic stress (Calbreath, 1992; Roychowdhury et al., 2024). Elevated LDH has been reported under various stress conditions, including nitrite exposure (Das et al., 2004; Jiang et al., 2024), endosulfan toxicity (Akhtar et al., 2010a; Kumar et al., 2025), starvation, thermal shifts (Roychowdhury et al., 2024), and crowding (Tejpal et al., 2009; Chatterjee et al., 2006). The present findings therefore imply that high TABP supplementation aggravated rather than mitigated stress. In contrast, the T2 group maintained LDH values close to those of the control, supporting the notion that this dose can help reduce infection-induced metabolic disturbance (Mohanty et al., 2023).

Collectively, the contrasting enzyme profiles of T2 and T3 indicate that TABP exerts a biphasic effect: at moderate levels it supports organ function and reduces leakage of tissue-derived enzymes into the bloodstream, but at higher inclusion rates it promotes abnormal elevation of diagnostic markers, signaling cellular damage, necrosis, and metabolic overload. This dose-dependent pattern highlights the need to carefully optimize the use of plant-derived feed additives in aquaculture. Based on the present results, TABP at the T2 level appears beneficial, whereas the higher dose (T3) is associated with adverse physiological consequences (Meena et al., 2024; Mohanty et al., 2023; Irshad et al., 2023). The integrated biomarker response (IBR) approach has further validated TABP as an effective immunomodulator in fish, illustrating that comprehensive biomarker strategies can serve as reliable tools for evaluating multiple physiological drivers in aquaculture nutrition research (Beliaeva and Siwicki, 2023).

Enzyme of oxidative damage

Aquatic animals confront oxidative stress mainly by reinforcing their antioxidant defence system, which represents the primary line of protection against cellular damage caused by reactive oxygen species (ROS) (Hermes-Lima, 2004). Among the key antioxidant enzymes, superoxide dismutase (SOD) converts superoxide radicals into hydrogen peroxide and oxygen (Villafranca et al., 1974), while catalase (CAT) further breaks down hydrogen peroxide into water and oxygen (Aebi, 1984). Glutathione-S-transferase (GST) complements these reactions by catalyzing the conjugation of reduced glutathione with toxic electrophiles, thereby limiting peroxidative damage (Hayes et al., 2005). Collectively, these enzymes are widely accepted as biomarkers of the oxidative status in fish exposed to both environmental and pathogenic stressors (Lushchak, 2011; Livingstone, 2020).

In this study, activities of SOD, CAT, and GST declined significantly up to the T2 group and increased markedly in T3. This pattern suggests that fish in T2 experienced a relatively mild oxidative challenge, so there was less demand for strong induction of defensive enzymes. In contrast, T3 fish appeared to face greater oxidative stress during infection, prompting a clear upregulation of their antioxidant machinery. The elevated antioxidant enzyme levels therefore reflect, on one side, the contribution of T. arjuna bark powder (TABP) bioactives in stimulating antioxidant pathways, and on the other, the fact that the higher inclusion level imposed a heavier oxidative load. This is consistent with earlier observations that modest phytobiotic supplementation tends to reduce oxidative pressure, whereas higher doses often trigger compensatory enzyme activation and can shift the system toward a pro-oxidant state (Spanou et al., 2011; Zhang et al., 2016; Wang et al., 2022; Nunes, 2024).

Recent work in aquaculture supports this interpretation. Abdel-Tawwab et al. (2018) reported increased SOD and CAT activities together with reduced lipid peroxidation in Nile tilapia fed polyphenol-rich diets under infection stress, while Liu et al. (2020) showed that herbal antioxidants helped carp maintain GST activity under ammonia exposure. Wang et al. (2022) demonstrated that flavonoid supplementation in fish diets enhanced antioxidant defences but that excessive intake could disturb redox balance. Dawood et al. (2020) and El Basuini et al. (2020) confirmed that dietary herbal extracts modulate antioxidant enzyme activities in fish under a range of stress conditions, and Abdel-Tawwab et al. (2024) found that Chlorella vulgaris nanoparticles, at optimal doses, elevated SOD and CAT activities but higher doses caused metabolic strain in Nile tilapia under hypoxia and cold stress. In the present case, the moderate diet (T2) appears to have minimized oxidative stress efficiently, whereas the higher level (T3) elicited strong enzymatic responses at the expense of greater metabolic burden.

Taken together, these findings indicate that TABP influenced redox homeostasis in a dose-dependent manner. At moderate supplementation, fish coped more effectively, with reduced induction of stress-related enzymes, whereas higher levels activated a robust antioxidant response that signifies both enhanced enzymatic protection and a heightened oxidative challenge. This clearly underscores the importance of dosage optimization to harness the protective effects of T. arjuna bark powder while avoiding the risk of provoking pro-oxidant conditions (Ahmadifar et al., 2021b; Mohanty et al., 2023).

Enzyme of neurotransmission (acetyl choline esterase)

Acetylcholinesterase (AChE) is a key enzyme in the nervous system, where it breaks down acetylcholine to regulate the transmission of nerve signals between cells. When AChE activity changes, it is often interpreted as an indicator of neurophysiological stress in fish. In our study, feeding L. rohita with 15 g/kg of T. arjuna bark powder (TABP) increased AChE activity, suggesting that the higher dose influenced the regulation of this neural enzyme. Similar context-dependent shifts have been reported previously; Akhtar et al. (2012), for example, observed lower AChE activity in rohu fingerlings held at 33 °C compared with those at normal rearing temperatures. In earlier work, environmental toxins and chemical contaminants such as organophosphate pesticides and a range of pollutants are known to depress AChE activity (Sancho et al., 1997; Dembele et al., 2000; Akhtar et al., 2010b). The usual interpretation is that acetylcholine accumulates at synapses when its breakdown is impaired. What stood out in our trial was that the moderate dose (T2) suppressed AChE activity, but this suppression was not seen at the higher dose (T3), where AChE levels rose instead. This pattern resembles responses described with other dietary additives. For instance, tryptophan supplementation prevented the decline of AChE in mrigal under crowding stress (Tejpal et al., 2009) and showed a similar stabilizing effect in rohu exposed to nitrite stress (Ciji et al., 2012). It appears that T. arjuna at higher inclusion levels can modulate AChE activity in a manner comparable to these nutraceutical strategies. The literature also reports anti-cholinesterase properties in various herbal preparations (Balkrishna et al., 2019; Suganthy et al., 2018), which supports the idea that bioactive compounds in T. arjuna contributed to the enzyme changes we recorded. Recent studies have shown that gold nanoparticles derived from T. arjuna bark extract exhibit marked anticholinesterase and neuroprotective effects (Suganthy et al., 2018; Manna, 2024), and water-based extracts of Terminalia species display notable anti-AChE potential in experimental setups (Balkrishna et al., 2019).

Under normal circumstances, a drop in AChE activity leads to increased synaptic acetylcholine, which can stimulate higher catecholamine release and promote glycogen breakdown, potentially elevating blood glucose. Interestingly, in our experiment blood glucose levels did not rise in any treatment group; they remained stable over the 90-day feeding period, the subsequent 60-day pond phase, and the 30-day intraperitoneal inoculation trial. This glucose stability occurring alongside the AChE changes suggests that the dietary treatment modified neural enzyme regulation without triggering major downstream metabolic disturbances.

Thus, our results indicate that TABP at elevated concentrations can adjust acetylcholinesterase activity in L. rohita without producing noticeable metabolic or behavioral side effects. While this raises the possibility that plant-derived bioactive compounds can influence neural function, it also highlights the importance of careful dosing, since altering cholinesterase activity too strongly in either direction may pose neurophysiological risks. Future studies that monitor neurotransmitter levels and behavior-linked endpoints would help clarify what these biochemical shifts mean for the wellbeing and performance of the fish (Behairy et al., 2023).

Digestive enzyme activity

Digestive enzymes play a central role in nutrient utilization and growth in fish because they break down dietary macromolecules into smaller, absorbable units. Earlier studies have established that enzyme activity in larvae is closely tied to the development of the digestive tract, while in adults, diet and feeding habits exert a stronger influence. Much of the available information still focuses on larval ontogeny, often in carnivorous marine species, due to the difficulty of formulating early-stage artificial diets. Data on freshwater carps, especially regarding the effect of plant-based additives, remain relatively limited. More recent work on Cyprinus carpio and Oreochromis niloticus has shown that herbal-based functional feeds can enhance digestive enzyme profiles and improve growth, further underlining the need for carp-specific studies (Goswami et al., 2020; Firmino et al., 2021; Kuebutornye et al., 2024).

In the present work with L. rohita, amylase activity showed no marked differences among the treatments during indoor feeding trials, indicating that carbohydrate digestion was not strongly affected by dietary T. arjuna bark powder (TABP). In contrast, studies on other carps supplemented with Azadirachta indica and Ocimum sanctum have reported variable but sometimes elevated amylase responses, suggesting species- and additive-specific modulation. Protease activity, however, was enhanced in the T2 group, and alkaline phosphatase (ALP) followed the same trend, pointing to improved protein digestion and absorptive processes at this supplementation level. Recent reports in Clarias gariepinus fed Moringa oleifera-based additives describe a similar protease-driven improvement in protein digestion and growth (Dawood et al., 2021; Elabd et al., 2019). Following bacterial challenge, the same general pattern was observed, although Edwardsiella tarda produced a stronger suppressive effect on enzyme activity than Aeromonas hydrophila, consistent with earlier findings that virulent bacterial strains can differentially impair digestive physiology. Interestingly, earlier work showed that amylase activity does not always correlate with dietary protein intake, which may help explain the inconsistent amylase responses seen here. In contrast, the enhanced protease activity observed with TABP supplementation likely reflects improved dietary protein utilization in juveniles, in line with data from other phyto-additive trials in warm-water fish (Ahmadifar et al., 2021a).

Under pond conditions, enzyme activities varied little between the T2 and T3 groups, probably because natural food sources supplied additional enzymes or substrates that masked dietary effects. In intraperitoneal inoculation experiments, protease, ALP, and lipase increased slightly in T3 at 15 days, but this trend did not persist. By 30 days, enzyme activities declined in all treatments except the control, indicating that infection may stimulate a short-term enhancement of digestive function, followed by a return to basal levels once immune and metabolic demands stabilize. Field-based work on Channa striatus exposed to bacterial challenge has similarly reported transient elevation of digestive enzymes early in infection, followed by decline as disease progresses (Rasal et al., 2023).

Previous studies confirm that herbal extracts can positively affect digestive physiology in fish. Sahu et al. (2007) reported beneficial effects of plant additives on digestive enzymes in L. rohita. More recent studies in other aquaculture species show a similar dose-dependent response: enzyme activities rise with moderate supplementation but decline once inclusion exceeds the optimal range. In the present investigation, protease activity was strongly correlated with weight gain (R ² = 0.95), which confirms its pivotal role in growth performance, as also reported in diet-enzyme correlation studies on Labeo calbasu and Catla catla. TABP-fed groups had significantly higher protease activity than controls, possibly due to improved protein utilization from enhanced digestion. Lipase, by contrast, was largely unaffected in indoor or pond trials, though it increased temporarily under inoculation trials. Given its important role in lipid digestion, this suggests that TABP has only a limited influence on lipid metabolism under normal feeding conditions, an observation echoed in trials with T. arjuna based feeds in tilapia where lipid-related enzymes did not respond as strongly as proteolytic ones (Meena et al., 2024).

The findings support the view that TABP improves nutrient digestion in L. rohita primarily by stimulating protease and ALP activities, which in turn are closely linked to protein assimilation and growth. Its effect on carbohydrate and lipid-digesting enzymes appeared limited and context-specific, in line with emerging evidence that plant-based additives have a more pronounced effect on protein-utilizing pathways than on carbohydrate or lipid metabolism in herbivorous and omnivorous fish (Goswami et al., 2020; Kuebutornye et al., 2024). Importantly, the study also reinforces the idea that the benefits of plant additives are dose-dependent, with moderate supplementation offering the most consistent improvement in digestive physiology and growth efficiency, a principle increasingly highlighted in recent phyto-additive meta-analyses and review work in aquaculture nutrition (Dawood et al., 2021; Roychowdhury et al., 2024).

T3, T4, cortisol and glucose

Cortisol and glucose are among the most widely used indicators of stress in teleost fish, as they reflect activation of the hypothalamic–pituitary–interrenal (HPI) axis and the mobilization of energy reserves when conditions become challenging. Elevated glucose in circulation usually arises from glycogen breakdown and gluconeogenesis from amino acids, providing a rapid energy supply to support the fight-or-flight response. Due to its conserved structure and central role across vertebrates, cortisol has long served as a key marker for evaluating stress responses in fish (Hontela, 2005). Its main actions include helping to maintain osmotic balance and regulating energy homeostasis, both of which are crucial for successful adaptation during stress.

In this study, fish in the indoor feeding trial that received the moderate TABP dose (T2) had lower cortisol levels than the control group, which showed relatively higher concentrations of both cortisol and glucose. This indicates that T. arjuna bark powder, when supplied at an appropriate level, can temper HPI axis activity and help reduce stress. In contrast, the higher TABP dose (T3) did not show this protective effect and was linked to poorer growth and feed utilization, suggesting that over-supplementation may carry adverse consequences. In the outdoor pond trial, no significant differences in cortisol or glucose were observed between T2 and T3, most likely because natural food sources and broader environmental variability diluted the influence of the dietary treatment.

Catecholamines also rise under stress, and their stimulation of glycogen breakdown contributes to increased blood glucose. Any dietary strategy that can blunt this response is usually regarded as beneficial. In the present work, TABP supplementation at the optimum dose reduced stress-related glucose elevations, in line with earlier reports where specific dietary interventions helped control stress-induced hyperglycemia in carps (Tejpal et al., 2009; Kumar, 2008; Neeraj et al., 2011). Similar protective effects of plant extracts have been described in mammals; for example, T. arjuna bark extract lowered physiological stress markers in rats (Bhattacharjee et al., 2019). Recent studies show that plant-based extracts such as grape seed extract can effectively reduce stress-hormone disturbances and support endocrine stability in fish exposed to environmental pollutants (Korkmaz et al., 2025). In addition, essential oils have shown potential as sedative-like agents to alleviate vaccination-induced stress in farmed fish, lowering cortisol and glucose responses (Priya et al., 2024).

Thyroid hormones provided further insight into metabolic regulation. Fish in the T3 group showed lower T3 and T4 levels compared with T2, where thyroid hormone concentrations remained higher. This pattern agrees with earlier observations that stress in fish is often linked to reduced thyroid activity (Klaren et al., 2007; Deane and Woo, 2007). Because thyroid hormones influence growth, development, and metabolic rate (Peter, 2007), the higher growth rates recorded in T2 align with elevated T4, while the weaker growth in T3 corresponds with depressed thyroid hormone levels. These findings support the view that moderate TABP supplementation can help maintain endocrine balance and support growth, whereas higher doses may impair thyroid function and thereby reduce performance. Recent work confirms that thyroid hormones play a central role in regulating gene expression and a range of biological processes essential for growth and reproduction across vertebrates, with ecological implications for environmental adaptation (Jann et al., 2024).

The results suggest that T. arjuna supplementation affects stress physiology in a dose-dependent way. At moderate levels it tends to suppress cortisol and glucose increases, stabilize thyroid hormone profiles, and support growth. At higher doses, however, the beneficial effects fade and clear physiological disruptions emerge. These outcomes emphasize the need to carefully optimize the inclusion of plant-based additives in aquaculture, because the dose ultimately decides whether the response is adaptive or harmful. Recent comprehensive reviews highlight that neuroendocrine systems in fish, including the HPI axis, react to various stressors with distinct timing and intensity, and that plant-derived compounds can modulate these responses in a controlled and often beneficial manner (Yuan et al., 2025).

IBR analysis

The Integrated Biomarker Response (IBR) analysis offers a detailed, multi-layered view of the physiological and biochemical changes in L. rohita after dietary supplementation with T. arjuna bark powder (TABP) and subsequent exposure to A. hydrophila (AH) and E. tarda (ET). Rather than treating each biomarker as an isolated event, the IBR framework treats them as interconnected components of a broader physiological network, which allows for a deeper interpretation of how diet and infection jointly influence fish health. The radar-plot visualization of 30 serum biomarkers—including ALP, AMY, BIT, GLU, LDH, GGT, and the transaminases—reveals that TABP does not act on a single pathway but instead induces coordinated shifts across metabolic, hepatic, and immune-related functions. In this sense, the IBR approach moves beyond simple enzyme-by-enzyme reporting and instead highlights the system-level reorganization of physiology under combined nutritional and pathogenic stress.

One of the most striking patterns was the strong increase in AMY, BIT, GLU, and LDH in the T2 and T3 groups under both AH and ET challenges. This pattern suggests that the fish are shifting toward greater mobilization of carbohydrate-derived energy, altered handling of bilirubin, and enhanced glycolytic and anaerobic activity to meet the energetic demands of the immune response. Such changes align with studies showing that immune activation in fish is metabolically demanding and often accompanied by elevated glucose and related enzymes (Reverter et al., 2020; Corraze and Panserat, 2011). The rise in BIT and LDH further indicates that the liver is under functional load, managing bilirubin metabolism while sustaining higher rates of substrate turnover. During the 90-day feeding phase, the marked activation of BIT, GGT, and SGOT in T2 and T3, even without infection, implies that the diet itself induces a low-grade, adaptive modulation of hepatic and metabolic activity. This could be interpreted as a mild “pre-conditioning” effect, where the liver adjusts to the presence of bioactive compounds in TABP, thereby preparing the fish for a more efficient response when infection is later imposed.

Other enzymes such as ALP, AMY, CRT, GLU, and LIP showed moderate but consistent changes, which is typical of an adaptive metabolic state rather than frank pathology. The relatively stable profiles of CKNAC, SGPTD, and triglycerides in the non-challenged phase suggest that the system is able to maintain homeostasis at the cellular and metabolic level until the additional stress of infection is introduced. Once bacteria are encountered, however, serum enzyme activities increase sharply—particularly in AMY, BIT, GGT, LDH, and SGPT—mainly in the T2 and T3 groups. This pattern indicates that TABP-supplemented fish mount a more robust immunometabolic response, with the liver and related tissues actively participating in substrate redistribution, inflammatory signaling, and tissue repair. The parallel increases in lipase and protease during both the feeding and infection phases reinforce the idea that TABP enhances digestive efficiency, thereby supporting the energetic and anabolic demands of stress and immunity (Sahu et al., 2007; Sapre et al., 2010). The fluctuating ALP activity across the trial further underscores its dual role—not only in brush-border digestion and membrane-related processes but also in aspects of immune signaling and inflammatory responses. Repeated shifts in GGT, LDH, and bilirubin across phases confirm that these markers are highly responsive to both nutritional and immune stimuli, making them useful indicators of functional status rather than mere signs of damage (Vinagre et al., 2012; Vazzana et al., 2017).

From an endocrine standpoint, the IBR-based hormonal profile shows that cortisol and thyroxine (T4) are among the most responsive variables to both TABP inclusion and bacterial challenge, with T2 and T3 groups showing higher values after infection. The rise in cortisol is consistent with activation of the hypothalamic–pituitary–interrenal (HPI) axis, which regulates energy availability, osmoregulation, and immune modulation during stress (Barton, 2002; Bonga, 1997). In practical terms, this means that the supplemented fish are mounting a strong, integrated stress response rather than simply tolerating the insult passively. The increase in T4 further points to an upregulation of metabolic rate and immune activity, in line with earlier work linking thyroid hormones to growth, energy turnover, and stress adaptation in fish (Power et al., 2001; McCann et al., 2014). The relative stability of T3 under these conditions reflects its known resistance to short-term stress, suggesting that the system is adjusting primarily at the T4 level to meet immediate energetic and immunological demands without destabilizing the more tightly regulated T3 pool (Deane and Woo, 2007). Recent applications of the IBR approach to pollutant and thermal-stress models demonstrate that such multi-biomarker frameworks can capture the timing and hierarchy of hormonal and enzymatic responses, providing a temporal map of how neuroendocrine systems orchestrate adaptation (Heredia-García et al., 2022.

At the antioxidant level, the IBR data highlight catalase (CAT) as the most responsive enzyme, with marked elevation across both dietary and infection phases. This underlines CAT’s central role in decomposing hydrogen peroxide and thereby limiting oxidative damage, especially under conditions of increased metabolic flux and immune activation (Lushchak, 2011). Superoxide dismutase (SOD) shows moderate induction, reflecting ongoing but controlled detoxification of superoxide radicals, while the relatively stable activity of glutathione S-transferase (GST) suggests that the present stress regime does not heavily engage GST-mediated conjugation pathways under these experimental conditions (Lushchak, 2011). The IBR index has been used effectively to summarize multi-tissue oxidative responses, with, for example, cardiac tissue showing a particularly strong antioxidant capacity compared with the kidney under thermal stress, emphasizing that different organs contribute unequally to the overall redox balance (Schleger et al., 2024).

The neurophysiological component of the IBR assessment, reflected in acetylcholinesterase (AChE) activity, reveals that TABP-supplemented fish, especially in the higher-dose groups, undergo selective modulation of cholinergic function during ET challenge. This pattern suggests that the nervous system participates in the stress and immune response not only through classical neuroendocrine pathways (HPI axis) but also through more direct modulation of neurotransmitter metabolism. Recent work indicates that AChE serves as a sensitive biomarker of neurotoxicity in fish, where its inhibition is often linked to exposure to organophosphates and related neurotoxicants, whereas its adaptive upregulation can reflect compensatory adjustments under physiological stress (Stott and Hogstrand, 2021; Piszczek et al., 2017; Formicki et al., 2025). In the present context, the expanded AChE response in high-TABP groups under ET infection may therefore represent a neuroprotective adjustment aimed at preserving synaptic integrity in the face of immune and metabolic perturbation.

The IBR-driven analysis suggests that TABP supplementation, at moderate to high inclusion levels, induces a coordinated enhancement of digestive, metabolic, immunological, endocrine, and antioxidant systems in L. rohita. Rather than isolated improvements, the biomarker responses point to a system-wide realignment toward greater digestive efficiency, metabolic resilience, and immune readiness, all of which likely contribute to enhanced disease tolerance under bacterial challenge. The IBR framework not only confirms the functional efficacy of TABP but also provides a mechanistic backdrop for how plant-derived feed additives can be used to optimize fish health in aquaculture. Recent reviews and meta-analyses support that IBR-based evaluation offers a robust way to gauge the performance of phytogenic immunostimulants, with serum biomarkers serving as sensitive, integrative indicators of infection, stress, and dietary impact (Meena et al., 2024; Dogan, 2025). This moves the discussion from simply “whether” TABP works to “how” and “under which conditions” it exerts its most beneficial effects, ultimately guiding more practical recommendations for dose and application in commercial fish production.

Phytochemical architecture and statistical validation of Terminalia arjuna bark extracts for nutraceutical standardization

The present study establishes a comprehensive phytochemical and statistical rationale for the deployment of T. arjuna bark powder (TABP) as a standardized nutraceutical feed additive in aquaculture. LC-FTMS profiling resolved the bark metabolome into functionally distinct classes dominated by phenolic acids (gallic acid, ellagic acid), flavan-3-ols ((−)-epigallocatechin, epicatechin, catechin, epigallocatechin gallate), flavonols (quercetin, quercetin-3β-D-glucoside, myricetin, kaempferol), and signature pentacyclic triterpenoids (arjunolic acid, ursolic acid, 18-β-glycyrrhetinic acid). These constituents collectively orchestrate a multi-tiered antioxidant defense network that operates through direct free-radical scavenging, transition-metal chelation, and transcriptional upregulation of phase-II detoxification enzymes via the Nrf2-ARE pathway (Ahmadifar et al., 2021a; Awad and Awaad, 2017). The identification of arjunolic acid as a bark-specific marker corroborates earlier Ayurvedic and phytochemical reports that T. arjuna bark is uniquely enriched in cardioprotective and hepatoprotective triterpenes (Semwal et al., 2016; Mandal et al., 2013), while the co-occurrence of high-abundance ellagic acid and gallic acid provides the molecular basis for the exceptional ferric-reducing and radical-scavenging capacities observed in polar bark extracts.

The statistical architecture of the antioxidant indices reveals critical insights for nutraceutical standardization. Pearson correlation analysis demonstrated that total phenolic content (TPC) is the principal driver of radical-scavenging activity, explaining 68.9% of DPPH variance, while hierarchical regression confirmed that flavonoid content (FLV) is psychometrically redundant with TPC in the bark matrix (VIF = 19.6, negligible ΔR²). This redundancy arises because the bark flavonoid pool comprising epigallocatechin gallate, catechin, and quercetin glycosides constitutes the dominant phenolic fraction. Consequently, TPC alone can serve as a cost-effective, high-throughput quality-control marker for commercial TABP batches, eliminating the need for dual phenolic-flavonoid testing without sacrificing predictive accuracy (Meena et al., 2021a). In contrast, FRAP exhibited acceptable independence from TPC (VIF = 1.66), indexing a distinct electron-transfer chemistry that mirrors the glutathione-GPx recycling mechanism in vivo. We therefore propose a composite nutraceutical index combining TPC (as the proxy for hydrogen-donation and scavenging capacity) and FRAP (as the proxy for cellular reducing power) to most robustly predict the capacity of a given TABP formulation to enhance the full spectrum of fish enzymatic antioxidants.

The mechanistic link between these in vitro indices and in vivo enzymatic antioxidant defenses is well established. Dietary polyphenols and triterpenoids from TABP activate the Nrf2-ARE signaling cascade, transcriptionally inducing superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPx), and glutathione-S-transferase (GST) in fish hepatocytes and erythrocytes (Harikrishnan et al., 2012; Tarnecki et al., 2019). The near-perfect correlation between DPPH and ABTS (r = 0.991) and between DPPH and NO-scavenging (r = 0.990) indicates that these assays capture a unified dimension of ROS/RNS neutralization capacity. When administered as a feed additive, this pooled scavenging potential may reduce the oxidative burden on farmed fish under intensive culture stress—characterized by high stocking density, hypoxia, handling, and pathogen challenge—thereby sparing endogenous enzymatic reserves and preventing lipid peroxidation and protein carbonylation. The strong TPC-DPPH association (r = 0.830) further validates that standardized phenolic loads can reliably predict the magnitude of enzymatic antioxidant enhancement, positioning TPC as an actionable benchmark for feed formulation.

Beyond redox homeostasis, the identified phytoconstituents exert direct immunomodulatory and antimicrobial actions that underpin disease resistance. Epigallocatechin gallate and ellagic acid suppress NF-κB and COX-2 expression, down-regulating pro-inflammatory cytokines (IL-1β, TNF-α) while enhancing macrophage respiratory burst, lysozyme activity, and complement C3 titers (Awad and Awaad, 2017; Ahmadifar et al., 2021a). Arjunolic acid and ursolic acid contribute hepatoprotective and anti-steatotic effects that preserve liver architecture during bacterial challenge, while 18-β-glycyrrhetinic acid modulates the hypothalamic-pituitary-interrenal axis to prevent stress-induced depletion of CAT and GPx (Meena et al., 2022a; Mohanty et al., 2023). These bioactivities converge to improve post-challenge survival against A. hydrophila and E. tarda—the two most devastating Gram-negative pathogens in Indian carp aquaculture—suggesting that TABP may serve as a promising antibiotic-sparing strategy (Meena et al., 2022b; Meena et al., 2024). The anti-parasitic efficacy of quinine and betulin against Argulus bengalensis and Ichthyophthirius multifiliis, as reported in the broader literature, further broadens the theoretical prophylactic scope of TABP, addressing multiple etiological agents in integrated culture systems (Dawood et al., 2021).

The dose-response optimization of TABP is critical to harnessing these benefits while avoiding pro-oxidant or hepatotoxic risks at excessive inclusion levels. Our previous work demonstrated that 1% dietary TABP significantly enhanced immune-metabolic enzyme activities, improved protease-driven protein utilization (strongly correlated with growth performance; r = 0.974), reduced cortisol stress responses, and elevated thyroxine levels in L. rohita, whereas 1.5% inclusion induced hepatic stress characterized by elevated serum transaminases and endocrine disruption (Meena et al., 2024). The present phytochemical and statistical framework provides the analytical basis for this dose-dependency: higher inclusion levels may overwhelm phase-II detoxification capacity, leading to compensatory antioxidant activation and metabolic overload. By standardizing TABP batches to a minimum TPC benchmark (e.g., ≥400 µg GAE/mL extractive concentrate) and monitoring FRAP as a secondary index, feed manufacturers can ensure that each inclusion delivers immunomodulatory potency within the safe therapeutic window, thereby optimizing growth, enzymatic defense, and survival simultaneously.

From a sustainability perspective, the integration of standardized TABP into commercial aquafeeds offers a promising avenue to reduce reliance on prophylactic antibiotics in intensive aquaculture—practice that has been linked to antimicrobial resistance (AMR), antibiotic residues, and environmental contamination (Dada, 2015). However, it is important to emphasize that the present study did not directly measure antibiotic residues, AMR gene prevalence, pathogen burden in culture effluent, or conduct side-by-side comparisons with antibiotic-treated groups. Consequently, TABP should be regarded at this stage as a potential antibiotic-sparing strategy rather than a direct antibiotic substitute. The World Health Organization and Codex Alimentarius have increasingly emphasized the need for responsible seafood production; antibiotic residues in edible fish tissues not only raise trade-standard concerns but may also contribute to horizontal gene transfer of resistance determinants between aquatic and human pathogens (Meena et al., 2022b). If developed as an antibiotic-sparing additive, TABP might offer non-specific immune priming through the upregulation of SOD, CAT, and GPx without the persistent chemical footprint associated with synthetic antibiotics, though this hypothesis requires empirical validation in residue-monitoring and AMR-surveillance studies. Moreover, given its high phenolic and flavonoid content, TABP holds theoretical potential to retard lipid oxidation in post-harvest fish muscle, which could translate into preserved sensory attributes, color stability, and textural integrity during refrigerated storage benefits that warrant direct investigation in future shelf-life and sensory-evaluation studies (Meena et al., 2021a; Awad and Awaad, 2017).

The chemometric and multivariate approaches employed here encompassing correlation matrices, collinearity diagnostics, and hierarchical regression provide a promising analytical framework for future nutraceutical quality-control validation. Within the present dataset, TPC and FRAP emerged as non-redundant markers capable of capturing distinct dimensions of antioxidant capacity, and their statistical independence suggests they could serve as complementary benchmarks for batch characterization. However, it is important to acknowledge that these indices were derived from the bark material used in this study and were not validated against several independent TABP batches with corresponding in-vivo enzymatic responses. Consequently, the proposed composite index should be viewed as a candidate protocol awaiting multi-batch validation rather than a finalized commercial standard. Should future studies confirm that TPC and FRAP reliably predict in-vivo antioxidant enhancement across diverse production batches, this framework could offer a cost-effective, high-throughput alternative to redundant phenolic-flavonoid testing. Such standardization would ultimately be essential for commercial scalability, because without validated analytical benchmarks, herbal products remain vulnerable to compositional drift that can undermine efficacy and safety. Until direct batch-to-batch validation is completed, we present this chemometric architecture as a theoretically grounded starting point for developing robust QC markers in phytogenic feed additives.

Conclusion

This study establishes T. arjuna bark powder (TABP) as a standardized, multi-functional phytogenic nutraceutical capable of enhancing enzymatic antioxidant defenses and disease resistance in farmed fish through a defined phytochemical–physiological pathway. LC-FTMS profiling and chemometric analyses resolved the bark metabolome into distinct functional classes dominated by phenolic acids (gallic acid, ellagic acid), flavan-3-ols ((−)-epigallocatechin, epicatechin, catechin, epigallocatechin gallate), flavonols (quercetin, myricetin, kaempferol), and signature pentacyclic triterpenoids (arjunolic acid, ursolic acid, 18-β-glycyrrhetinic acid), with polar bark extracts exhibiting the highest in vitro bioactivity. Statistical validation identified total phenolic content (TPC) as the principal, non-redundant predictor of radical-scavenging capacity, while ferric-reducing antioxidant power (FRAP) indexed a complementary electron-transfer mechanism that mirrors glutathione-dependent recycling in vivo. Together, these markers provide a robust, cost-effective quality-control framework that eliminates redundant flavonoid and duplicate scavenging assays, enabling batch-to-batch consistency in commercial TABP formulations.

In juvenile L. rohita, dietary supplementation with 1% TABP for 90 days optimized growth performance and digestive physiology, attenuated cortisol-mediated stress responses, elevated thyroxine levels, and markedly improved survival following challenge with Aeromonas hydrophila and Edwardsiella tarda. Integrated Biomarker Response analysis confirmed that this moderate dose induced a coordinated, system-wide physiological realignment toward metabolic resilience and immune readiness, reflecting efficient redox homeostasis without excessive oxidative burden. In contrast, 1.5% inclusion provoked hepatocellular damage, compensatory antioxidant enzyme activation signaling oxidative overload, and endocrine disruption, underscoring the critical importance of precise dose optimization.

Collectively, these findings demonstrate that TABP operates through a dual-mode mechanism: its bioactive constituents directly neutralize reactive oxygen and nitrogen species in vitro, while optimized dietary administration transcriptionally and enzymatically fortifies endogenous antioxidant defenses in vivo. By translating phytochemical and statistical benchmarks into antibiotic-free feed formulations, standardized TABP integration offers a scalable strategy to reduce antimicrobial resistance, eliminate chemical residues in edible tissues, mitigate environmental contamination, and enhance post-harvest seafood quality and shelf-life through muscle lipid antioxidant protection. This positions TABP as a cost-effective, eco-friendly alternative to synthetic antibiotic growth promoters, advancing the sustainable intensification of aquaculture in alignment with global One Health objectives for fish welfare, environmental stewardship, and consumer protection.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Sabina Passamonti, University of Trieste, Italy

Reviewed by: Antonia Mataragka, Agricultural University of Athens, Greece

Ntsoaki Malebo, Central University of Technology, South Africa