Effects of different dietary starch sources on growth performance, hepatic antioxidant capacity, and muscle quality of adult grass carp (Ctenopharyngodon idellus)

TL;DR

This study shows that using sorghum as a starch source in grass carp feed improves growth, liver antioxidants, and muscle quality compared to other starches.

Contribution

The study identifies sorghum as a sustainable and effective alternative to wheat in grass carp feed.

Findings

Sorghum improved final body weight and feed conversion ratio in grass carp.

Sorghum enhanced hepatic antioxidant enzyme activities and gene expression.

Sorghum improved muscle quality by increasing flavor-related amino acids and fatty acids.

Abstract

Starch is a key component of herbivorous fish feeds, but its utilization efficiency varies across species. This study examines the effects of replacing wheat with different starch sources (barley, corn, sorghum, and broken rice) on the growth performance, hepatic antioxidant capacity, and muscle quality of grass carp (Ctenopharyngodon idellus), aiming to identify suitable starch sources for grow-out feed and provide a theoretical basis for developing efficient, environmentally friendly, and safe feed. Sorghum was chosen due to its superior nutritional profile, strong antioxidant properties, and potential as a sustainable alternative to wheat. The results showed that final body weight (FBW) was significantly higher in the S1 groups than in S2 (P < 0.05), the Feed Conversion Ratio (FCR) of the S2 group was higher than that of the S1 group. Biochemical analyses revealed that triglyceride (TG) and aspartate aminotransferase (AST) activities were similar and higher in S1 and S4 than in other groups. Hepatic superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), and catalase (CAT) activities in S4 were comparable to S1 and significantly higher than in S2 and S5 groups. Additionally, malondialdehyde (MDA) levels were significantly higher in S2 than in S1, S3, and S4. At the molecular level, S4 showed significantly higher expression of hepatic glutathione peroxidase (gpx), cu/zn-superoxide dismutase (cuznsod), nuclear factor E2-related factor 2 (nrf2), and heme oxygenase 1 (ho-1), while the expression of Kelch-like ECH-associated protein-1 (keap1a) was downregulated. Sorghum promoted the accumulation of flavor-related amino acids and beneficial fatty acids in muscle tissue. In conclusion, compared to barley, corn, and broken rice, sorghum is an ideal and sustainable alternative to wheat in grass carp feed, offering significant potential for improving aquaculture feed formulations.

•Compared to other starch sources, sorghum improved grass carp growth•Sorghum enhanced SOD, GSH-Px, and CAT activity in grass carp over other starch.•Sorghum enhanced gpx, cuznsod, nrf2, and ho-1 expression compared to other starch.•Compared to other starch, sorghum improved the muscle quality of grass carp

Compared to other starch sources, sorghum improved grass carp growth

Sorghum enhanced SOD, GSH-Px, and CAT activity in grass carp over other starch.

Sorghum enhanced gpx, cuznsod, nrf2, and ho-1 expression compared to other starch.

Compared to other starch, sorghum improved the muscle quality of grass carp

Figures

The relative mRNA expression levels of hepatic antioxidant-related genes in *Ctenopharyngodon idellus*.Notes: Values are presented as mean ± SEM (*n* = 3). Gene expression was normalised using the 2^-ΔΔCt^ method with β-actin as the benchmark. Data are presented as relative expression levels. Mean values for the same gene with different letters were significantly different (*P* ˂ 0.05). *gpx,* glutathione peroxidase*; cuznsod,* copper‑zinc superoxide dismutase*; nrf2,* nuclear factor-E2 related factor2*; keap 1a,* kelch-like ECH-associated protein-1*; ho-1,* heme oxygenase 1*.*Fig. 1

The liver hematoxylin-eosin staining (400 ×, scale bars indicate 50 μm) of grass carp in the S1 (A), S2 (B), S3 (C), S4 (D), and S5 (E), respectively. The sections were cut using a microtome (Leica RM2235，Leica Biosystems, Nussloch, Germany) provided by Seville Biotechnology Co., Ltd., with a thickness of 5 μm. The staining was performed using the standard Hematoxylin and Eosin (H&E) staining method.Fig. 2

Full text

Introduction

1

With the sustainable intensification of aquaculture, optimizing feed formulations and reducing dependence on marine ingredients has become a key measure to improve the production efficiency of herbivorous species, especially species like grass carp (Ctenopharyngodon idellus), which, due to their exceptional carbohydrate utilization capacity, play a significant role in the efficient use of plant-based ingredients and the sustainable utilization of feed resources (Zhi, 2024; Wang et al., 2015; Kamalam et al., 2017). In the diet of grass carp, starch is particularly important as an energy source because it is a cost-effective alternative to lipids and proteins (Wilson et al., 1994; Ahmed, Naeem and Liu, 2022, Li et al., 2020). However, the nutritional and metabolic value of starch varies depending on its origin, structure, and physicochemical properties (Zhai et al., 2024). Therefore, identifying suitable feed alternatives is a critical breakthrough in addressing the imbalance between feed supply and demand, and this issue remains a focus of current research in nutrition and feed science (Jannathulla et al., 2019; Serra et al., 2024).

In this experiment, wheat was replaced with four starch sources—barley, corn, sorghum, and broken rice—to investigate the role of starch origin as a key determinant of feed efficiency and physiological homeostasis. The utilization efficiency and physiological effects of different starch sources are influenced by factors such as the amylose-to-amylopectin ratio, particle size, and non-starch polysaccharide (NSP) concentration (Cai et al., 2018; Singh, Dartois, & Kaur, 2010). Rapidly digestible starches from wheat and corn are rapidly hydrolyzed, providing immediate energy but potentially inducing postprandial hyperglycemia and hepatic oxidative stress (Lygren & Hemre, 2001). In contrast, starches with higher amylose or resistant starch fractions, such as those from sorghum and barley, are digested more slowly, and this difference may help enhance antioxidant defense and metabolic stability (Huang et al., 2023; Tanwar et al., 2023; Irondi, Adewuyi, & Aroyehun, 2022). Similarly, rice-based by-products, such as broken rice, become a cost-effective feed option for herbivorous species due to their small, easily gelatinized starch granules and their contribution to circular bioeconomy goals (Huervana et al., 2024; Lin et al., 2022). These functional differences have been validated across multiple species. Given its strong carbohydrate tolerance and efficient carbohydrate metabolism, grass carp is well-suited for diets that include various starch sources. Recently, the utilization of extruded compound feed in grass carp farming has increased, improving overall farming efficiency. Grass carp play a significant role in aquaculture; however, studies on the effects of different starch sources on grass carp growth performance, hepatic antioxidant capacity, and muscle quality, particularly in large individuals, remain limited. Most existing research has focused on single starch sources, such as wheat and corn, and has not fully explored the potential of alternative sources like barley, sorghum, and broken rice (Chen et al., 2012; Wang et al., 2025). Building on the structure–function concept reported for natural polymers, including starch, where functional behavior is governed by intrinsic structural characteristics and material origin (Braşoveanu et al., 2023), subsequent research demonstrated that variations in starch botanical source and structural organization directly translate into distinct physicochemical and functional properties, underscoring the critical role of starch source selection in applied systems (Braşoveanu et al., 2024; Wong et al., 2009). Differences in carbohydrate metabolism between starch sources may influence grass carp growth and health, yet the mechanisms remain insufficiently explored. Despite grass carp's strong carbohydrate tolerance, their responses to various starch sources may differ due to variations in digestion and nutrient absorption.

The present study aimed to evaluate the effects of replacing wheat with different starch sources (barley, corn, sorghum, and broken rice) on growth performance, hepatic antioxidant capacity, and muscle quality in adult grass carp fed isonitrogenous and isolipidic diets for 16 weeks. The findings are intended to provide a scientific basis for selecting appropriate starch sources in efficient feed formulation.

Materials and methods

2

For all procedures relating to live animals, approval has been authorized by the Institutional Animal Care and Use Committee of the Guangdong Ocean University (ID GDOU-IACUC-2021-A0207).

Experimental diets

2.1

The composition and formulation of the experimental diets are presented in Table 1. Wheat, barley, corn, sorghum, and broken rice were used as the starch sources to formulate five isonitrogenous and isolipidic diets, designated as S1, S2, S3, S4, and S5. Each diet contained 36.5% fat and 6.8% protein. The raw materials, finely ground through an 80-mesh sieve, were mixed for 15 min to ensure uniformity. Pure water was then added, and the mixture was stirred for 10 min to achieve complete homogenization. The homogenized mixture was extruded using a twin-screw extruder (XBF-62, Xinbei Fa, Shandong, China), producing pellets with a diameter of 4.5 mm. The extrusion parameters were set as follows: barrel temperatures were maintained at 90 °C, screw speed at 150 rpm, and feed rate at 30 kg/h. The moisture content of the mixture before extrusion was 30%. The extruded pellets were immediately transferred to a drying machine at a temperature of 60 °C for 4 h. After drying, the pellets were coated with soybean oil (refined) using a vacuum spraying system (XBF-65, Xinbei Fa, Shandong, China) at the inclusion rate shown in Table 1. An antioxidant mixture (ethoxyquin + BHT, 7:3 ratio, 100 mg/kg oil) was added to the soybean oil in proportion. Once spraying was completed, the pellets were allowed to cool to room temperature (25 ± 2 °C). Finally, the pellets were sealed in moisture-resistant aluminum foil pouches and stored at −20 °C in the dark until use in a deep freezer (DW-40L328, Haier, Qingdao, China).Table 1. Composition and nutrient levels of the experimental diets (dry matter basis, %).Table 1. IngredientsGroupS1S2S3S4S5Wheat26.00.00.00.00.0Barley0.026.00.00.00.0Corn0.00.026.00.00.0Sorghum0.00.00.026.00.0Broken rice0.00.00.00.026.0Soybean oil4.13.83.53.54.2Calcium dihydrogen phosphate2.02.02.02.02.0Choline chloride0.10.10.10.10.1Rapeseed meal30.830.830.830.830.8Premix^a^1.01.01.01.01.0Rice bran meal11.010.09.610.59.0Dehulled soybean meal25.025.025.025.025.0Wheat gluten0.01.32.01.11.9Total100.0100.0100.0100.0100.0Nutritional levelCrude protein32.333.032.832.332.6Crude lipid7.56.76.86.66.8Notes:1 ^a^ Premixes provided by Guangdong Daynew Aquatic Technology Co: vitamin A, 250000 IU/kg; vitamin D^3^, 135000–160,000 IU/kg; d1-alpha tocopherol acetate, 2.1 g/kg; vitamin K3, 0.24 g/kg; vitamin B1, 0.67 g/kg; and vitamin B2, 1.35 g/kg. Vitamin B6, 0.5 g/kg; Vitamin B12, 1.35 mg/kg; Calcium D-Pantothenate, 1.8 g/kg; Niacinamide, 3.0 g/kg; Folic Acid, 0.21 g/kg; D-Biotin, 4.0 mg/kg; Inositol, 6.0 g/kg; l-Ascorbic Acid-2-Phosphate, 10.5 g/kg; Fe, 9.5–25.0 g/kg; Cu, 0.4–1.0 g/kg; Zn, 2.0–6.0 g/kg; Mn, 1.25–5.0 g/kg; I, 00.1–1.0 g/kg; Se, 0.01–0.04 g/kg; Co, 0.125–0.18 g/kg. Nutrients levels were calculated values. Values after the decimal point are reported in the form of significant figures.

Experimental animal and feeding management

2.2

The research was carried out at the aquaculture facility of Guangdong Dening Aquatic Technology Co., Ltd., situated in Xingtan Town, Foshan, Guangdong Province, China. Grass carp juveniles were procured by a local fish breeding facility in Sanshui District, Foshan (Guangdong, China), and were initially placed in a holding pond for 2 W to acclimatize and meet the experimental requirements. Sex was not determined due to the juvenile stage of the fish. After the acclimation period, the fish were fasted for one day. The fish were disinfected using Povidone Iodine Solution. A total of 375 juvenile grass carp (392.13 ± 0.62 g) were randomly allocated to 15 net cages (2.5 × 2.5 × 2.5 m^3^), with 25 fish per cage. Sample size was based on previous grass carp nutrition studies using similar designs and was considered sufficient to detect biologically relevant differences between dietary treatments. Random allocation was performed using a random number table to assign fish to cages, minimizing initial bias and ensuring balance between experimental groups. The experimental design included five dietary treatments (S1–S5), each with three replicates, and 25 fish per replicate to ensure the repeatability and reliability of the experiment. Dietary treatments were randomly assigned to cages, and fish selected for sampling were chosen using the same randomization procedure. The experimental unit was the cage (n = 3 per dietary treatment). Individual fish served as observational units. The daily feeding and data collection were recorded by cage number, with cage numbers not corresponding to the experimental group order. The feeding staff were unaware of the experimental group assignments. Fish were cultured for 16 weeks and fed ad libitum at 08:00, 12:00, and 17:00 each day. Throughout the research, all groups were kept under natural light–dark cycles. The aquatic temperature fluctuated ranging from 21.3 °C to 33.5 °C, ammonia nitrogen concentration levels were kept between 0.05 and 0.10 mg/L, concentration of nitrites ranged from 0.05 to 0.15 mg/L, and the average dissolved oxygen concentration was maintained greater than 5.00 mg/L. Regular checks on water quality parameters (pH, ammonia nitrogen, nitrite, and dissolved oxygen) were conducted to ensure optimal environmental conditions. Health monitoring of the grass carp was performed, including regular observations for abnormal behavior or signs of illness. Immediate intervention was provided if any issues were detected. If mild imbalance, sluggish swimming, reduced feeding, abnormal respiration, localized fin rot, minor bleeding, or lethargy are observed, immediately isolate the affected fish for observation. If no improvement occurs within 24 h, perform humane euthanasia.

Sample collection

2.3

Sample collection and analysis was conducted blind to dietary treatment.

Following the feeding trial, grass carp were fasted for 24 h prior to sampling. Fish were randomly selected from each cage for sampling, with three fish per cage chosen using the same randomization process. According to standard protocols for deep anesthesia, the fish were anesthetized with MS-222 (Guangzhou Kangxin Pharmaceutical Co., Ltd., Huangpu District, Guangzhou, China) at a concentration of 200 mg/L, buffered with sodium bicarbonate to a pH of 7.0–7.5. Once deep anesthesia was achieved, euthanasia was performed using an overdose of the anesthetic solution at a concentration of 300 mg/L. The following measurements were recorded: FBW, WG, SGR, FCR, FI, VSI, HIS, and CF.

For liver tissue collection, three fish were randomly picked, and their liver samples were quickly dissected. Two pieces of uncompressed liver tissue from each fish (0.5 × 0.5 × 0.5 cm^3^) were placed into 5 mL centrifuge tubes containing 4% paraformaldehyde solution, ensuring that the samples were fully submerged, for histological (H&E) sectioning. Similarly, liver tissue from another three fish was minced (roughly the size of a soya), immersed in RNA later (Ambion, Thermo Fisher Scientific, Waltham, MA, USA) and kept at −80 °C for analysis of the target gene mRNA expression.

For enzyme activity analysis, the livers and hindguts of three additional fish were preserved in liquid nitrogen for subsequent enzyme activity analysis.

For muscle tissue analysis, dorsal muscle tissues from three selected fish were kept at −20 °C for routine, amino acid, and fatty acid analysis. Three additional fish were selected, and complete back muscle samples were cut into rectangular pieces (112 cm^3^), placed on ice, and analyzed within 24 h for texture characteristics with a texture measurement device (TMS-PRO, Sterling, Virginia, USA) based on the texture profile analysis (TPA) method.

Proximate composition analysis

2.4

The basic composition of the experimental diets, as well as the whole-body and muscle tissue composition of grass carp, was analyzed using methods recommended by the Association of Official Analytical Chemists (AOAC, 2005). Moisture content (MS) was quantified by drying the samples in an oven at 105 °C until no further weight loss was observed. The sample weights before and after drying were recorded to determine the moisture content. Crude protein (CP) content was analyzed by Kjeldahl's method using the Haineng K9860 automatic Kjeldahl nitrogen analyzer (Haineng, Shandong, China). Crude lipid (CL) was extracted using a Soxhlet apparatus with petroleum ether (boiling range: 30–60 °C) as the solvent, with an extraction time of 8 h. Crude ash (CA) was determined by incinerating carbonized samples in a muffle furnace at 550 °C for 4 h, with a sample size of 3 g (Wei et al., 2022).

Enzyme activity measurement

2.5

Serum biochemical indices, including glucose (GLU No. A154-1-1), triacylglycerol (TG No. A110-1-1), total cholesterol (TC No. A111-1-1), alanine aminotransferase (ALT No. C009–2-1), and aspartate aminotransferase (AST No. C010–2-1), were quantified.

Hepatic antioxidant enzyme activities, namely superoxide dismutase (SOD No. A001-3-2), glutathione peroxidase (GSH-Px No. A005-1-2), catalase (CAT No. A007-1-1), and malondialdehyde (MDA No. A003-1-2), were also assessed.

Digestive enzyme activities, including α-amylase (AMS No. C016-1-2), lipase (Lip No. A054-2-1), and protease (Pro No. A080-2-2), were measured as well.

All biochemical indicators were examined with commercial test kits (Jiancheng Bioengineering Institute in Nanjing, China) as per the manufacturers' protocols.

Extraction of RNA from liver tissue and quantitative real-time PCR analysis

2.6

Liver RNA was extracted using the TransZol Up Plus RNA Kit (No. ER501-01, TransGen Biotech Co., Ltd., Beijing, China), and its integrity was assessed via 1% agarose gel electrophoresis. RNA extraction, quality assessment, and cDNA synthesis were conducted according to the method described by Li et al. (2022). The RT-qPCR procedure and reaction program were carried out according to the protocols outlined by Xu et al. (2021) and Abouel Azm et al. (2022). The RNA concentration and purity were determined using a spectrophotometer. Following extraction, the RNA was reverse-transcribed to cDNA using the PrimeScript RT-PCR Kit (Accurate Biotechnology (Hunan) Co., Ltd., ChangSha, China). The target gene sequences are listed in Table 2. RT-qPCR was conducted using a LightCycler® 480 (Roche Diagnostics, Mannheim, Germany) with β-actin as the reference gene. The reaction mixture (10 μL) included 5 μL of 2× SYBR® Green Pro Taq HS Premix II (Accurate Biotechnology Hunan Co., Ltd.,China), 3.2 μL nuclease-free water, and 1 μL of cDNA with 0.4 μL each of the forward and reverse primers. The relative expression of the target gene was calculated using the 2^-△△Ct^ method (Livak & Schmittgen, 2001).Table 2. Nucleotide sequences of the primers used to assay gene expressions by qPCR.Table 2. Target geneForward primer (5′–3′)Reverse primer (5′–3′)Sourcesβ-actinTATGTTGGTGACGAGGCTCAGCAGCTCGTTGTAGAAGGTGAbouel Azm et al., 2022gpxTGCAACCAGTTCGGACATCACATCAGGGACACAGCGTCATXu et al., 2021cuznsodCGCACTTCAACCCTTACAACTTTCCTCATTGCCTCCXu et al., 2021nrf2GGAGAAGAGCGAACGTAGCAGGAACGAGAAAAACGGTGCCXu et al., 2021keap 1aCTGCGAAAGCGAGGTCTACATTGCCTTTGGAGGAACGTCGXu et al., 2021ho-1GCTGAGATACGCCCAGAGACCAGCTCGATGCTGTTCATGCXu et al., 2021Notes: cuznsod, cu/zn-superoxide dismutase, gpx, glutathione peroxidase; nrf2, nuclear factor-E2 related factor2; keap 1a, kelch-like ECH-associated protein-1; ho-1, hemeoxyenase-1.

Growth performance calculations

2.7

Fish were fed three times daily at 08:00, 12:00, and 17:00. Feed intake is controlled by providing feed based on the consumption observed in preliminary trials. After 60 min, any uneaten feed is removed and recorded, and the amount for the next feeding is adjusted to prevent overfeeding. This procedure was standardized across all groups to ensure comparability.

The following parameters were measured by independent researchers who were blinded to the experimental groups during the measurement process. The data collected during the measurement were not linked to the experimental group information until the completion of the experiment.

The formula is as follows:

$$[eqn]$$ $$[eqn]$$ $$[eqn]$$ $$[eqn]$$ $$[eqn]$$ $$[eqn]$$ $$[eqn]$$

Note: m₁, m₂: Initial and final average body weight of fish (g). n₁, n₂: Initial and final number of fish. t: Experimental period (d). I: Total dry feed intake (g). m₃, m₄: Initial and final total body weight of fish (g). m₅: Liver weight (g). m₆: Individual fish weight (g). m₇: Visceral weight of individual fish (g). L: Total body length of individual fish (cm).

Statistical analysis

2.8

Blinding was used throughout the data collection and analysis process to ensure objectivity and minimize bias. The data are presented as means ± standard error of the mean (SEM) and analyzed with SPSS version 27.0 (SPSS Inc., Chicago, IL, USA) using one-way ANOVA, followed by Tukey's honestly significant difference test. The experimental unit was the cage, with three replicates per dietary treatment (n = 3). In the trial, the indicator analysis is exploratory. The Tukey HSD test was chosen for its suitability in multiple comparisons with homogeneous variance, providing better statistical power, especially with large sample sizes. Statistical significance was considered at P < 0.05.

Result

3

Growth performance

3.1

For FBW, the S1 group was significantly higher than the S2 group (P < 0.05; Table 3). The FCR of the S2 group was higher than that of the S1 group (P < 0.05). No significant differences were observed for other indicators.Table 3. Effects of different starch sources on growth performance for Ctenopharyngodon idellus.Table 3. ItemsGroupsS1S2S3S4S5FBW(g)1643.9 ± 30.9^a^1525.7 ± 36.2^b^1609.3 ± 38.1^ab^1623.9 ± 16.9^ab^1554.3 ± 67.6^ab^WG (%)316.8 ± 9.5289.1 ± 13.1312.9 ± 10.3314.9 ± 7.9297.2 ± 17.4SGR(%/d)1.3 ± 0.021.3 ± 0.031.3 ± 0.021.3 ± 0.021.3 ± 0.04FCR1.7 ± 0.1^b^1.9 ± 0.1^a^1.8 ± 0.1^ab^1.8 ± 0.1^ab^1.9 ± 0.1^ab^SR (%)100.0 ± 0.0100.0 ± 0.0100.0 ± 0.0100.0 ± 0.0100.0 ± 0.0VSI (%)10.1 ± 0.610.1 ± 0.110.6 ± 0.410.3 ± 0.111.1 ± 0.5HSI (%)2.2 ± 0.22.1 ± 0.12.1 ± 0.22.1 ± 0.13.8 ± 1.7CF(g/cm^3^)2.0 ± 0.02.0 ± 0.02.1 ± 0.12.1 ± 0.12.0 ± 0.1Data are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05). Values after the decimal point are reported in the form of significant figures.

Whole-body and muscle compositions

3.2

There was no significant effect on the whole-body and dorsal muscle MS, CP, CL, or CA contents of grass carp (P > 0.05; Table 4 and Table 5). Although the muscle CL content was higher in the S2 group and lower in the S1 group, there was no statistically significant difference.Table 4. Effects of different starch sources on whole-body for Ctenopharyngodon idellus (dry matter basis, %).Table 4. ItemsGroupsS1S2S3S4S5Moisture content69.2 ± 0.368.8 ± 0.368.5 ± 0.168.3 ± 0.172.3 ± 3.6Crude protein49.1 ± 0.149.1 ± 0.048.6 ± 0.248.7 ± 0.248.8 ± 0.2Crude lipid28.9 ± 0.128.6 ± 0.228.2 ± 0.131.9 ± 3.428.6 ± 0.1Crude ash12.3 ± 0.212.4 ± 0.112.2 ± 0.312.5 ± 0.312.1 ± 0.3Data are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05). Values after the decimal point are reported in the form of significant figures.Table 5. Effects of different starch sources on muscle composition for Ctenopharyngodon idellus (dry matter basis, %).Table 5. ItemsGroupsS1S2S3S4S5Moisture content72.6 ± 0.472.4 ± 0.672.5 ± 0.572.8 ± 0.172.2 ± 0.2Crude protein74.8 ± 0.0374.8 ± 0.075.0 ± 0.175.0 ± 0.174.8 ± 0.0Crude lipid7.1 ± 0.48.3 ± 0.28.1 ± 0.68.2 ± 0.98.1 ± 0.5Crude ash5.5 ± 0.15.5 ± 0.15.5 ± 0.05.5 ± 0.15.5 ± 0.1Data are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05). Values after the decimal point are reported in the form of significant figures.

Serum biochemical indices

3.3

Dietary starch sources significantly affected biochemical serum markers in grass carp (P < 0.05; Table 6). GLU levels were significantly lower in the S4 group compared to S1, S2, S3, and S5 (P < 0.05). TG concentration in the S2 group was significantly lower than in S1 and S4 (P < 0.05). AST activity in the S4 group was significantly higher than in S1 and S3 (P < 0.05). ALT activity was significantly higher in S2 compared to all other groups (P < 0.05).Table 6. Effects of different starch sources on serum biochemical indices for Ctenopharyngodon idellus.Table 6. ItemsGroupsS1S2S3S4S5GLU (mmol/L)2.6 ± 0.0^b^2.6 ± 0.0^b^3.3 ± 0.1^a^2.4 ± 0.1^c^3.4 ± 0.0^a^TG (mmol/L)3.4 ± 0.0^a^2.6 ± 0.1^c^3.0 ± 0.1^b^3.2 ± 0.1^ab^3.0 ± 0.0^b^AST (U/L)7.1 ± 0.0^b^7.0 ± 0.0^ab^6.7 ± 0.2^b^7.2 ± 0.0^a^7.1 ± 0.0^ab^ALT (U/L)8.5 ± 0.0^b^9.1 ± 0.1^a^8.5 ± 0.0^b^8.4 ± 0.0^b^8.5 ± 0.0^b^Data are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05). Values after the decimal point are reported in the form of significant figures.

Hepatic antioxidant metabolism

3.4

Dietary starch sources significantly affected the hepatic antioxidant-related gene expression in grass carp (P < 0.05; Fig. 1). S4 group exhibited markedly increased mRNA expression of gpx, cuznsod, nrf2, and ho-1, while keap1a expression was markedly lower than in other groups (P < 0.05). Among the groups, S4 group showed the highest expression of cuznsod and ho-1, whereas the lowest expression levels were found in the S3 group. In contrast, keap1a showed the lowest expression in S4.Fig. 1. The relative mRNA expression levels of hepatic antioxidant-related genes in Ctenopharyngodon idellus.Notes: Values are presented as mean ± SEM (n = 3). Gene expression was normalised using the 2^-ΔΔCt^ method with β-actin as the benchmark. Data are presented as relative expression levels. Mean values for the same gene with different letters were significantly different (P ˂ 0.05). gpx, glutathione peroxidase*; cuznsod,* copper‑zinc superoxide dismutase*; nrf2,* nuclear factor-E2 related factor2*; keap 1a,* kelch-like ECH-associated protein-1*; ho-1,* heme oxygenase 1*.*Fig. 1

Antioxidant capacity

3.5

Dietary starch sources had a significant impact on the hepatic antioxidant capacity of grass carp (P < 0.05; Table 7). S1 group presented the greatest SOD activity, which was markedly greater than that of remaining groups (P < 0.05). S4 group also showed significantly higher SOD activity than S2 and S5 groups (P < 0.05), whereas S2 group had the minimal activity. Regarding GSH-Px activity, S1 group again exhibited the highest value, substantially exceeding those of S2, S3, S4, and S5 groups (P < 0.05). S4 group had markedly higher GSH-Px activity than S2, S3, and S5 groups (P < 0.05), with the lowest level observed in S2. For CAT activity, S4 groups showed significantly higher values than S2, S3, and S5 groups (P < 0.05). Conversely, MDA followed an opposing pattern. S2 group had markedly greater MDA levels than S1, S3, and S4 groups (P < 0.05) (P < 0.05).Table 7. Effects of different starch sources on the hepatic antioxidant capacity for Ctenopharyngodon idellus.Table 7. ItemsGroupsS1S2S3S4S5SOD (U/mg prot)345.3 ± 3.8^a^211.7 ± 10.3^d^274.1 ± 1.2^bc^286.0 ± 2.5^b^259.1 ± 3.4^c^GSH-PX (U/mg prot)104.4 ± 1.1^a^86.3 ± 0.8^c^91.5 ± 1.9b^c^98.7 ± 0.5^a^93.0 ± 1.3^b^CAT (U/mg prot)20.3 ± 0.1^ac^19.4 ± 0.1^c^19.5 ± 0.2^c^20.8 ± 0.1^a^19.6 ± 0.2^bc^MDA (nmol/mg prot)3.02 ± 0.2^d^5.7 ± 0.1^a^3.5 ± 0.4^cd^4.4 ± 0.05^bc^5.2 ± 0.1^ab^Data are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05). Values after the decimal point are reported in the form of significant figures.

Digestive enzyme activities

3.6

Starch sources markedly affected activities of digestive enzymes for grass carp (P < 0.05; Table 8). Por, S1 group was substantially greater than that in S2, S3, and S5, but not significantly different from S4 (P < 0.05). Lip, S3 group was substantially greater than that in S2, S4, and S5 groups (P < 0.05). AMS, S5 group displayed significantly higher values than all remaining groups (P < 0.05).Table 8. Effects of different starch sources on intestinal digestive enzyme activities for Ctenopharyngodon idellus.Table 8. ItemsGroupsS1S2S3S4S5Pro (U/mg prot)196.3 ± 9.5^a^1544.8 ± 21.3^d^1826.9 ± 20.7^b^1949.3 ± 11.8^a^1735.5 ± 15.9^c^Lip (U/g prot)73.9 ± 0.2^a^71.5 ± 0.1^d^72.9 ± 0.1^b^73.3 ± 0.2^ab^72.2 ± 0.1^c^AMS (U/mg prot)5.5 ± 0.0^b^4.2 ± 0.1^d^4.7 ± 0.0^c^5.1 ± 0.1^c^6.2 ± 0.0^a^Data are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05). Values after the decimal point are reported in the form of significant figures.

Hepatic morphology

3.7

H&E staining of liver sections showed normal hepatic structures in all treatment groups, except for the S2 group (Fig. 2).Fig. 2. The liver hematoxylin-eosin staining (400 ×, scale bars indicate 50 μm) of grass carp in the S1 (A), S2 (B), S3 (C), S4 (D), and S5 (E), respectively. The sections were cut using a microtome (Leica RM2235，Leica Biosystems, Nussloch, Germany) provided by Seville Biotechnology Co., Ltd., with a thickness of 5 μm. The staining was performed using the standard Hematoxylin and Eosin (H&E) staining method.Fig. 2

Texture properties

3.8

Although no significant differences were observed among groups (P < 0.05; Table 9), variations in muscle texture indices were noted. Hardness was highest in S1, followed by S5 and S4, with S2 showing the lowest value. Adhesiveness was negative in all groups, with the lowest value in S3. Springiness was slightly higher in S3 and S4, while cohesiveness remained similar across all groups. Gumminess and chewiness were highest in S1 and lowest in S2, while resilience was consistent across groups.Table 9. Effects of different starch sources on dorsal tissue characteristics for Ctenopharyngodon idellus.Table 9. ItemsGroupsS1S2S3S4S5Hardness2383.8 ± 285.22109.4 ± 330.72215.3 ± 37.12264.3 ± 190.42247.5 ± 122.8Adhesiveness−1.5 ± 1.0−1.0 ± 0.6−0.8 ± 0.3−1.3 ± 0.4−1.3 ± 0.6Springiness0.5 ± 0.00.5 ± 0.00.5 ± 0.00.5 ± 0.00.4 ± 0.0Cohesiveness0.3 ± 0.00.3 ± 0.00.3 ± 0.00.3 ± 0.00.3 ± 0.0Gumminess726.9 ± 35.7616.9 ± 57.2664.8 ± 43.1648.8 ± 8.4606.7 ± 47.8Chewiness337.8 ± 14.1273.9 ± 17.7340.6 ± 43.1327.9 ± 13.2267.9 ± 34.2Resilience0.2 ± 0.00.2 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.0Data are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05). Values after the decimal point are reported in the form of significant figures.

Amino acid composition

3.9

Notable differences (P < 0.05; Table 10) were observed in several amino acids between treatment groups. Threonine and serine contents were the greatest in the S4 group. Glycine concentration was notably greater in the S4 group compared with S1, S2, and S5 groups. Tyrosine content in S4 was markedly greater than S1. Proline levels in S1 were substantially higher than in S2(P < 0.05). The total delicious amino acids (∑DAA) were significantly elevated in S4 relative to S2. For other amino acids, total amino acids (∑AA), and essential amino acids (∑EAA), there were no notable differences among (P > 0.05).Table 10. Effects of different starch sources on sorsal muscle amino acid composition (g/100 g, as dry basis) for Ctenopharyngodon idellus.Table 10. ItemsGroupsS1S2S3S4S5Aspartic acid8.7 ± 0.38.5 ± 0.08.9 ± 0.29.2 ± 0.18.7 ± 0.3Threonine3.7 ± 0.1^ab^3.6 ± 0.0^b^3.7 ± 0.0^ab^3.8 ± 0.0^a^3.7 ± 0.1^ab^Serine3.4 ± 0.1^ab^3.2 ± 0.1^b^3.4 ± 0.1^ab^3.5 ± 0.0^a^3.3 ± 0.1^ab^Glutamic acid12.9 ± 0.412.6 ± 0.113.0 ± 0.213.4 ± 0.112.7 ± 0.3Glycine4.1 ± 0.1^b^4.1 ± 0.0^b^4.3 ± 0.0^ab^4.5 ± 0.1^a^4.1 ± 0.1^b^Alanine4.8 ± 0.14.8 ± 0.04.9 ± 0.15.1 ± 0.14.8 ± 0.1Valine4.0 ± 0.24.1 ± 0.04.2 ± 0.14.2 ± 0.04.1 ± 0.1Methionine2.4 ± 0.12.4 ± 0.02.4 ± 0.02.5 ± 0.02.4 ± 0.0Isoleucine3.8 ± 0.23.8 ± 0.03.9 ± 0.13.9 ± 0.03.8 ± 0.1Leucine6.8 ± 0.26.8 ± 0.07.0 ± 0.17.1 ± 0.06.8 ± 0.1Tyrosine3.2 ± 0.2^b^3.3 ± 0.0^ab^3.4 ± 0.1^ab^3.6 ± 0.0^a^3.4 ± 0.1^ab^Phenylalanine3.5 ± 0.23.5 ± 0.03.6 ± 0.03.6 ± 0.03.5 ± 0.0Lysine7.9 ± 0.48.0 ± 0.08.2 ± 0.18.4 ± 0.18.0 ± 0.1Histidine3.1 ± 0.33.1 ± 0.03.4 ± 0.13.5 ± 0.03.3 ± 0.1Arginine5.1 ± 0.25.1 ± 0.05.2 ± 0.05.2 ± 0.05.0 ± 0.1Proline2.8 ± 0.0^a^2.7 ± 0.0^b^2.8 ± 0.1^a^2.8 ± 0.0^ab^2.8 ± 0.0^ab^∑AA80.3 ± 2.979.6 ± 0.282.4 ± 1.084.6 ± 0.680.4 ± 1.6∑EAA31.1 ± 1.431.3 ± 0.132.2 ± 0.532.9 ± 0.131.4 ± 0.4∑DAA40.0 ± 1.17^ab^39.1 ± 0.1^b^40.8 ± 0.5^ab^42.1 ± 0.4^a^39.7 ± 1.0^ab^Data are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05).∑AA = ∑Amino acids; ∑EAA = ∑Essential amino acids; ∑DAA = ∑Delicious amino acids. Significance probability associated with the F-statistic. Values after the decimal point are reported in the form of significant figures.

Fatty acid composition

3.10

Significant differences in fatty acid composition were detected among treatments (P < 0.05; Table 11). The total polyunsaturated fatty acids (ΣPUFA) were significantly greater in the S4 group compared with S5 group (P < 0.05), while no markedly differences were found among S1, S2, S3, and S4 groups. The total n-3 fatty acids (Σn3) in S4 were markedly greater than those in S2 and S5 groups. There were no notable differences among for other individual fatty acids (C14:0–C22:6n3) or total saturated fatty acids (ΣSFA) (P > 0.05).Table 11. Effects of different starch sources on dorsal muscle fatty acid composition (g/100 g, as dry basis) for Ctenopharyngodon idellus.Table 11. ItemsGroupsS1S2S3S4S5C14:01.6 ± 0.21.5 ± 0.11.4 ± 0.01.5 ± 0.11.6 ± 0.1C14:1n50.1 ± 0.00.1 ± 0.00.1 ± 0.000.1 ± 0.00.1 ± 0.0C15:00.1 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.0C16:022.6 ± 0.221.8 ± 0.522.2 ± 0.421.8 ± 0.421.7 ± 0.4C16:1n78.2 ± 0.27.3 ± 0.36.5 ± 0.48.0 ± 0.68.4 ± 0.9C17:00.1 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.0C17:1n70.2 ± 0.00.2 ± 0.00.2 ± 0.00.2 ± 0.00.2 ± 0.0C18:03.8 ± 0.24.0 ± 0.14.3 ± 0.23.8 ± 0.13.7 ± 0.1C18:1n9c37.3 ± 1.739.3 ± 0.636.9 ± 1.336.8 ± 1.539.3 ± 1.9C18:2n6c15.1 ± 1.415.1 ± 0.216.7 ± 1.015.9 ± 1.014.5 ± 1.8C20:00.1 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.0C18:3n60.2 ± 0.00.2 ± 0.00.3 ± 0.00.2 ± 0.00.2 ± 0.0C20:11.3 ± 0.01.3 ± 0.01.1 ± 0.01.1 ± 0.01.3 ± 0.0C18:3n31.2 ± 0.011.5 ± 0.11.6 ± 0.11.6 ± 0.11.3 ± 0.1C21:00.1 ± 0.00.1 ± 0.00.2 ± 0.00.1 ± 0.00.1 ± 0.0C20:21.5 ± 0.01.6 ± 0.01.6 ± 0.11.6 ± 0.11.6 ± 0.1C20:3n61.1 ± 0.11.0 ± 0.01.2 ± 0.11.1 ± 0.11.0 ± 0.1C20:3n30.1 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.00.1 ± 0.0C20:4n63.4 ± 0.52.9 ± 0.13.7 ± 0.33.8 ± 0.63.1 ± 0.5C20:5n30.2 ± 0.00.2 ± 0.00.2 ± 0.00.2 ± 0.00.2 ± 0.0C22:6n31.3 ± 0.21.1 ± 0.01.3 ± 0.11.4 ± 0.21.1 ± 0.2∑SFA28.7 ± 0.228.0 ± 0.528.6 ± 0.527.7 ± 0.427.6 ± 0.4∑MUFA46.7 ± 2.5ᵃᵇ48.2 ± 0.9ᵃ44.8 ± 1.7ᵇ46.2 ± 2.1ᵃᵇ49.3 ± 2.7ᵃ∑PUFA24.6 ± 2.5ᵃᵇ23.8 ± 0.3ᵃᵇ26.6 ± 1.2ᵃ26.1 ± 1.7ᵃ23.1 ± 2.9ᵇ∑n33.1 ± 0.3ᵃ2.9 ± 0.1ᵃᵇ3.2 ± 0.2ᵃ3.4 ± 0.3ᵃ2.7 ± 0.3ᵇ∑n619.9 ± 2.1ᵃ19.3 ± 0.4ᵃ21.8 ± 1.1ᵃ21.2 ± 1.5ᵃ18.9 ± 2.5ᵃ∑n3/∑n60.2 ± 0.0ᵃᵇ0.1 ± 0.0ᵃᵇ0.1 ± 0.0ᵇ0.2 ± 0.0ᵃ0.1 ± 0.0ᵇPUFA/SFA0.9 ± 0.1ᵃᵇ0.8 ± 0.0ᵃᵇ0.9 ± 0.0ᵃ0.9 ± 0.0ᵃ0.8 ± 0.1ᵇData are mean ± SEM (n = 3). Different superscripts in the same column are significantly different (P < 0.05).SFA, saturated fatty acids; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; n3, Omega-3 Fatty Acids; n6, Omega-6 Fatty Acids. Values after the decimal point are reported in the form of significant figures.

Discussion

4

The capacity of aquatic species to metabolize carbohydrates is influenced by the type of starch. Utilizing different starch sources as energy can modify activities of digestive enzymes and the rate of glucose release, thereby leading to differences in growth performance among fish (Enes et al., 2012; Krogdahl et al., 2005; Azaza et al., 2020). In this investigation by Wang et al. (2025), no negative effects on the growth performance of grass carp were observed when dietary sorghum inclusion levels reached 20%. Lins Rodrigues et al. (2020) demonstrated that sorghum can effectively replace corn in fish feed and resulted good growth performance and physiological indices under phytase supplementation. The current study, different starch sources markedly affected the growth development of grass carp. When sorghum was used as the starch source, FBW and FCR were comparable to those of the S1 group and higher than in the other groups. No significant differences were observed in the other growth-related indices, indicating that starch source replacement did not affect these parameters, particularly sorghum, did not produce negative health effects or condition of the grass carp. Therefore, sorghum, as an effective alternative starch source to wheat, can effectively maintain the performance in growth and feed conversion efficiency of grass carp.

Starch is a key energy source in feed and plays an important role in regulating the nutritional composition of fish body and muscle (Wang et al., 2023; Ma et al., 2023). Body composition reflects the nutritional status and metabolic balance of fish, with water being the primary component. It not only provides the medium for metabolic reactions but also affects the transport and storage efficiency of nutrients within the body. The nutritional value of whole fish is predominantly determined by protein and fat content (Xiong et al., 2019). Typically, higher protein content enhances nutritional value, whereas excessive fat content can result in softer flesh and diminished flavor (Feng et al., 2024). This study revealed that different starch sources in the diet had no significant effect on MS, CP, CL, or CA content in both whole fish and muscle of grass carp. This phenomenon is consistent with recent reports: under isocaloric and isonitrogenous conditions, replacing carbohydrate or starch sources usually does not usually significantly alter the component composition of entire fish or muscle, only showing slight fluctuations (Ye et al., 2025). Additionally, another study indicated that moderate differences in starch sources have a limited impact on fat deposition in muscle (Cheng et al., 2025). Moreover, diets with well-matched protein and fat levels can further buffer the fluctuations in body composition caused by starch replacement through the “sparing effect” (He et al., 2024; Liao et al., 2025). Therefore, the substitution of wheat with sorghum as a source of starch performed well in maintaining the stability of grass carp body composition.

As a key indicator of fish metabolism, serum biochemical parameters reflect the overall nutritional status and metabolic condition, providing direct insights into how dietary composition, particularly energy sources, influence metabolic processes (Yuan & Tao, 2023). GLU, the principal substrate for energy metabolism in fish, is of great importance in osmotic regulation, energy supply, and glucose–lipid homeostasis (Lei et al., 2022; Deng et al., 2018; Liu et al., 2020). Excessively high GLU levels often indicate overly digestible carbohydrate sources that induce transient hyperglycemia and metabolic burden, whereas moderately reduced GLU levels generally reflect more efficient energy utilization and metabolic stability (Qian et al., 2024, Rasal et al., 2020). In the present study, the plasma GLU level in the S4 group was markedly lower than that in remaining groups, suggesting that sorghum starch releases glucose more slowly, thereby contributing to glycemic stability and reducing metabolic stress. Previous studies have similarly reported that starches processed through slow-release techniques improve glycemic control and insulin sensitivity in fish (Song et al., 2018). TG is another key indicator of lipid metabolism and liver condition. In the current study, TG level in the S4 group was comparable to that in the S1 group but higher than that in remaining groups. It has been documented that both the level and type of dietary carbohydrates can modulate hepatic lipogenic pathways and affect circulating lipid profiles (Lei et al., 2022; Liu et al., 2022). Additionally, polyphenols and resistant starch fractions from sorghum may further optimize glucose and lipid metabolism, while mitigating excessive lipid deposition (Collins et al., 2024; Balbuena-Pecino et al., 2024). Regarding hepatic enzyme activity, AST was highest in the S4 group, while ALT was highest in the S2 group. The increased ALT observed in S2 group may be associated with slower starch digestion and a higher metabolic load, suggesting a compensatory hepatic stress response (Liu et al., 2022). Overall, sorghum starch demonstrated advantages in maintaining glucose homeostasis and metabolic balance. The combination of lower GLU levels and moderate hepatic enzyme activities indicates more efficient energy utilization and reduced metabolic stress (Ridha et al., 2023). Therefore, substituting sorghum for wheat promotes serum biochemical stability and supports healthy growth.

As the liver is the central organ for nutrient metabolism and redox regulation in fish, its structural integrity and antioxidant defense system are critical for maintaining physiological homeostasis under different dietary conditions (Xie et al., 2024). Histological observation of liver sections stained with H&E revealed that all treatment groups exhibited normal hepatic structures without noticeable pathological alterations. Oxidative stress is a key factor affecting the redox homeostasis and overall health of fish, particularly under nutritional or metabolic stress conditions (Lushchak, 2011, Sun et al., 2025). The hepatic antioxidant system of fish includes key enzymes such as GPx, SOD, and CAT, which play central roles in eliminating reactive oxygen species (ROS) and maintaining cellular stability (Köroğlu et al., 2024; Martínez-Álvarez et al., 2005). The results of this study showed that the activities of SOD, GPx, and CAT in the S4 group were comparable to those in the S1 group. MDA exhibited an opposite trend. The type and digestibility of carbohydrates play a critical role in regulating oxidative stress (Lin et al., 2025). Easily digestible or high-starch diets tend to induce hyperglycemia and mitochondrial electron leakage, leading to excessive ROS production, whereas slow-release or resistant starches help stabilize energy supply and mitigate oxidative stress (Liang et al., 2022; Wang et al., 2023).

Mechanistically, various dietary factors can enhance antioxidant defenses through activation of nrf2 signaling pathway. Previous research in fish models have shown that polyphenols or antioxidant peptides can reduce MDA content and increase SOD and CAT activities (Seo et al., 2023; Xu et al., 2022; Zarei et al., 2022). Consistent with these findings, in the current study, the hepatic expression of gpx, cuznsod, nrf2, and ho-1 in the S4 group was significantly upregulated, while keap1a expression was lower than in the other groups, indicating transcriptional activation of the antioxidant defense system. The upregulation of nrf2 and its downstream target ho-1 suggests that sorghum starch may enhance the redox balance of hepatocytes by activating the nrf2–keap1 axis (Kobayashi and Yamamoto, 2005). Similarly, Zhong et al. (2022) found that resistant starch increased nrf2 and ho-1 expression and alleviated oxidative damage in tilapia, while Wang et al. (2023) reported that resistant starch improved the antioxidant capacity and lipid metabolism of fish. Sorghum starch has moderate gelatinization properties, likely due to its polyphenols and resistant starch components with antioxidant activity, which may regulate the transcription of nrf2-regulated genes (Safari et al., 2020, Wu et al., 2017). These active compounds can induce the nuclear translocation of nrf2 and upregulate antioxidant genes such as ho-1 and gpx, thereby enhancing ROS-scavenging capacity (Moratilla-Rivera et al., 2023). These findings collectively indicate that sorghum starch exerts multi-level antioxidant modulation. In contrast, barley and broken rice, which contain a higher proportion of digestible starch, tend to cause rapid increases in blood glucose and excessive mitochondrial ROS production, inducing oxidative stress and downregulating antioxidant gene expression (Ahmed et al., 2022; Huang et al., 2021). The elevated expression of gpx and cuznsod observed in the S4 group was consistent with their enzymatic activity trends, further confirming that sorghum starch enhances antioxidant responses at both transcriptional and enzymatic levels. Activation of the nrf2–keap1–ho-1 pathway is thus a key mechanism by which sorghum starch improves hepatic oxidative stability and reduces lipid peroxidation (Liu et al., 2020). Therefore, replacing wheat with sorghum starch upregulating the expression and activities of key antioxidant enzymes such as gpx and cuznsod.

As intestinal digestive enzymes are key determinants of nutrient utilization efficiency, their activities provide important insights into how different dietary starch sources influence digestive physiology in fish (De Almeida, Lundstedt and Moraes, 2006, Villasante et al., 2019, Lv et al., 2021). Intestinal digestive enzymes form a critical link between feed intake and nutrient utilization, hydrolyzing proteins, lipids, and carbohydrates into absorbable small molecules, thereby determining growth performance and feed efficiency (Navarro-Guillén et al., 2022). The activities of these enzymes also reflect the fish's adaptation to dietary composition and intestinal condition (Zheng et al., 2025; Liu et al., 2023). The current study revealed that different starch sources significantly affected the digestive enzyme activities in the intestines of grass carp. Regarding Pro, the sorghum group exhibited activity comparable to that of S1 group and higher than that of other groups, suggesting that both wheat and sorghum diets are beneficial for protein digestion. For Lip, S3 group showed the highest activity, while S1 and S4 groups also maintained relatively high levels with no significant difference between them, indicating that the digestibility of carbohydrates and the energy release patterns may regulate lipid digestion efficiency (Sriranjani et al., 2025; Zhou et al., 2022; Li et al., 2022). For AMS, S5 group showed the greatest activity, whereas S2 group showed the minimal, emphasizing how starch structure and gelatinization degree influence amylase secretion and activation (Cione et al., 2021). Dietary starch type and processing method strongly affect intestinal physiology and enzymatic responses: highly gelatinized and easily digestible starches generally enhance amylase activity and starch digestibility, whereas resistant starch fractions reduce proximal hydrolysis and are more likely to undergo fermentation in distal segments (Sriranjani et al., 2025, Maas, Verdegem, Wiegertjes and Schrama, 2020). Studies on tilapia have shown that diets with sorghum starch yield favorable apparent digestibility and enzymatic responses, supporting its feasibility as an alternative carbohydrate source (Bonvini et al., 2018). Furthermore, improving intestinal integrity or microbiota composition through optimized culture systems or functional nutritional supplementation can enhance trypsin, amylase, and lipase activities, thereby promoting nutrient absorption (Amenyogbe et al., 2024). In summary, both S1 and S4 groups exhibited clear advantages in protease activity (and partially in lipase), favoring protein and lipid utilization; S5 group promoted amylase activity, whereas S2 group, due to its high proportion of resistant starch, showed the lowest protease and amylase activities, consistent with limited proximal hydrolysis (García-Meilán et al., 2023). Therefore, sorghum, as an excellent alternative starch source to wheat, can optimize enzymatic responses and nutrient absorption while maintaining digestive stability. However, the digestive characteristics and energy conversion pathways of different starches may still fine-tune through the gut-liver axis, and the mechanisms need to be further validated through metabolic enzyme activity, energy allocation, and gut transcriptomics (Ma et al., 2023; Xiong et al., 2024).

Finally, muscle quality is a crucial factor determining the economic value and consumer acceptability of fish (Hu et al., 2025). The sensory characteristics of fish fillets—including texture, flavor, and nutritional composition—play a key role in consumer preference and market demand (Moreno et al., 2012; Li et al., 2023; Cheng et al., 2014). The results of this study suggest that although there were no notable differences among the treatment groups with different starch sources, the texture parameters of grass carp muscle still exhibited some variations. The hardness of S1 group was the highest, followed by S4 group and S5 group, while S2 group was the lowest. This indicates that the type of starch can, to a certain extent, affect the water-holding capacity and myofibrillar structure of fish muscle, as hardness and chewiness are closely related to myofibril density and collagen cross-linking (Ahmed et al., 2022). The physicochemical properties of dietary carbohydrates can affect the energy metabolism and nutrient deposition of fish muscle. In this study, replacing wheat with sorghum starch did not have adverse effects on texture parameters such as cohesiveness and resilience, indicating that replacing wheat with sorghum starch can maintain the integrity of muscle structure. Zhao et al. (2018) and Prakash et al. (2025) also pointed out that appropriate carbohydrate sources can enhance fish muscle hardness and elasticity by promoting muscle fiber development and collagen stability. Although the gender differences in grass carp or their specific impact on muscle growth have not been explicitly explored in this study, referencing the findings from Barbel Steed's research, it can be inferred that gender differences may, to some extent, influence the growth performance and muscle quality in grass carp (Li et al., 2023). Further research is needed to verify this hypothesis.

The amino acid composition plays a significant role in determining the nutritional value and flavor quality of fish. Amino acids such as threonine, serine, glycine, alanine, and glutamic acid are considered the main components contributing to the umami and sweetness of fish meat (Ma et al., 2020). In this research, the contents of threonine, serine, glycine, and tyrosine in S4 group were greater than other groups, and the total content of flavor amino acids was better than that in other groups. These outcomes indicate that sorghum starch will promote the synthesis or retention of free amino acids in muscle, thereby improving the flavor quality of fish meat. Because sorghum starch has moderate gelatinization characteristics and a balanced ratio of amylose to amylopectin, it can provide stable energy for protein metabolism, which is conducive to the deposition of amino acids in muscle (Kasumyan, 2016). The fatty acid composition of fish meat not only reflects its nutritional value but also reflects its health-promoting properties (Noreen et al., 2025). Polyunsaturated fatty acids, especially the n-3 series, are believed to enhance the nutritional quality of fish meat and have positive effects on cardiovascular health and anti-inflammatory functions in consumers (Mohammadzadeh & Feizy, 2025; Zhang et al., 2024).

The muscle texture, amino acid composition, and fatty acid composition of grass carp are interrelated. Although no significant direct effects were observed across the treatments, correlations between muscle hardness, chewiness, and the levels of amino acids and fatty acids were noted. Specifically, the S4 group had higher levels of glycine, proline, and tyrosine, which are key for collagen synthesis and stability, thereby influencing muscle hardness and elasticity (Zhang et al., 2025). Additionally, S4 group demonstrated higher levels of n-3 polyunsaturated fatty acids and total polyunsaturated fatty acids, which likely contributed to improved muscle texture, enhanced tenderness, and better elasticity (Xia et al., 2025; Cabanillas-Gámez et al., 2024). These impacts may also be linked to collagen stability, muscle hydration, and reduced oxidative stress. Overall, sorghum starch, through its impact on amino acid and fatty acid composition, Improved the hardness and elasticity of grass carp muscle while maintaining its structural integrity This suggests that sorghum starch can be a more effective carbohydrate source for grass carp compared to barley, corn, and broken rice, provides stable energy, promoting the deposition of amino acids and fatty acids within muscle tissue, thereby enhancing meat quality and nutritional value.

This study investigates the effects of various starch sources on grass carp growth, antioxidant capacity, and muscle mass, with primary outcomes focused on growth, antioxidant, and muscle indices, and secondary outcomes including serum markers, digestive enzyme activities, and liver histology, emphasizing comparisons between the sorghum group, wheat control, and other starch sources. However, certain limitations remain. The absence of starch digestibility measurements, gut microbiota analysis, and polyphenol quantification in the research constrains our understanding of the potential mechanisms and nutritional components associated with starch sources. Additionally, the starch sources used in this study were whole grains rather than purified starches, leading to differences in starch structure, non-starch polysaccharides, resistant starch, endogenous enzymes, and antinutritional factors, the effects of which on grass carp remain unverified. Future research should address these gaps to further elucidate the metabolic and physiological effects of starch sources in fish. Ultimately, investigating the bioavailability of polyphenols and their impact on fish health would strengthen the conclusions of this study.

Conclusion

5

In conclusion, under the conditions of this study, S4 group had superior performance in FBW, WG, SGR, and FCR, compared to the other groups. Additionally, the sorghum group exhibited better liver antioxidant enzyme activities (SOD, GSH-PX, CAT, MDA) and relative gene expression (gpx, cuznsod, nrf2, keap1a, ho-1). Sorghum increased amino acids and unsaturated fats, boosting muscle nutrition. Therefore, sorghum is the most suitable starch source to replace wheat in grass carp feed formulation based on various indices.

Authors contribution

Haojie Lu: conducted the study, analyzed the data; wrote the original draft. Qihui Yang: designed the study and edited the manuscript. Beiping Tan and Yuhang Chen: revised the manuscript. Junguang Yuan: assisted in the conduct of the study and analyzed the data. Lei Zhong: participated in the interpretation of the results. Jianwei Wu and Liyi Zhou: assisted in the conduct of the feeding trial. All authors read and approved the final version of the manuscript for submission.

CRediT authorship contribution statement

Haojie Lu: Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation. Yuhang Chen: Methodology, Conceptualization. Lei Zhong: Methodology, Conceptualization. Junguang Yuan: Supervision, Methodology, Conceptualization. Jianwei Wu: Visualization, Validation, Software. Liyi Zhou: Visualization, Validation, Investigation. Beiping Tan: Writing – review & editing, Conceptualization. Qihui Yang: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition.

Funding

This work was supported by the Guangxi Key Research and Development Program (AB2506910013), Guangdong-Hong Kong-Macao Greater Bay Area (Foshan) Advanced Manufacturing National Centre for Excellence in Engineering Innovation Joint Training Support Program (2023FCXM020).

Declaration of competing interest

The authors affirm that they have no financial conflicts of interest or personal relationships that could be perceived as influencing the research presented in this paper.