The Impact of Feeding Rates on the Growth, Stress Response, Antioxidant Capacity, and Immune Defense of Koi (Cyprinus carpio var. koi)

Abstract

The feeding rate is critical for fish growth and the feed conversion ratio (FCR). Prior research has primarily evaluated the optimal feeding rate by examining growth performance and the FCR. Given the high cost of koi as an ornamental fish, it is essential to consider the effects of feeding rates on its welfare and health. This study aimed to examine the impact of the feeding rate on the growth, stress response, antioxidant capacity, and immune defense of koi. A total of 240 fish, with an initial body weight of 10.02 ± 0.18 g, were randomly assigned to six groups and fed expanded diets at 1%, 2%, 3%, 4%, 5%, and 6% of body weight per day (BW/day) for a duration of eight weeks. The study results indicated that the daily weight gain of fish at 3–4% BW/day was significantly greater than that of the other groups. Fish in both high and low feeding rate groups exhibited reduced antioxidant capacity and heightened inflammatory response, as indicated by a notable decrease in catalase (CAT) and superoxide dismutase (SOD) activity, along with diminished expression of cat and sod, as well as a significant increase in the expression of tnf-α, il1, and il10. The elevated feeding rate did not enhance immune defense mechanisms. Furthermore, a low feeding rate resulted in heightened stress and reduced immunity in koi, as indicated by a significant reduction in plasma cortisol (COR), complement C3, complement C4, total protein (TP), albumin (ALB), and globulin (GLO) levels, along with decreased activities of lysozyme (LZM) and myeloperoxidase (MPO). The optimal feeding rate for enhancing non-specific immunity in koi is a daily intake of 3–4% of body weight on extruded feed. Our findings reveal that underfeeding led to diminished antioxidants and immune defenses in koi, while overfeeding also proved detrimental. An optimal feeding rate of 3–4% body weight per day is necessary to enhance growth and health in this species.

1. Introduction

In commercial aquaculture, feed costs represent 60–70% of total expenses [1,2]. The feeding regime is a critical factor influencing feed costs, as it significantly impacts fish growth and feed utilization [3]. While feed cost reduction remains a key objective, inappropriate feeding rates may inadvertently escalate long-term operational expenses [4]. For instance, overfeeding not only increases feed waste but also induces chronic stress and immunosuppression, elevating disease treatment costs and mortality rates. Conversely, underfeeding compromises immune defenses, exacerbating susceptibility to pathogens. Thus, optimizing feeding rates should balance economic efficiency with health preservation. Consequently, optimal feeding rates have been established for several prominent aquaculture species. However, these studies have primarily prioritized growth performance at the expense of assessing the long-term physiological effects of feeding rates on stress responses and immune function [5,6,7], particularly in ornamental fish such as koi. This gap is critical because impaired health in high-value ornamental fish translates directly into economic losses due to mortality and reduced marketability, unlike food fish, where growth efficiency dominates economic prioritization. In large-scale and standardized culture conditions, factors such as feed formulation, feed type, culture density, light intensity, water temperature, dissolved oxygen, pH, and ammonia nitrogen are meticulously controlled within the optimal range for the specific culture species. Additionally, feeding is performed using feeding machines at predetermined time intervals. Consequently, the feeding rate may represent a significant stressor for fish in contemporary large-scale and standardized culture conditions [8]. Research indicated that stress influences the immune process, with chronic and persistent stress typically resulting in an immunosuppressive effect [9]. Feeding is a behavior that occurs throughout the life cycle of fish, and inappropriate feeding rates can lead to chronic stress [10]. This chronic stress, in turn, suppresses the immune system, increasing the fish’s susceptibility to disease [11].

Immune defense is a nutrient-intensive physiological process [12], and the availability of adequate nutrients to immune cells is crucial for their proper functioning. Research on blunt snout bream (Megalobrama amblycephala) [11], starry flounder (Platichthys stellatus) [13], and chinook salmon (Oncorhynchus tshawytscha) [5] indicates that nutrient supplementation beyond optimal growth levels improves fish immune function and disease resistance. The nutritional levels of fish in artificial aquaculture environments are primarily influenced by feed intake, which is determined by the feeding rate [14]. Economic losses stemming from diminished immunity and disease resistance in modern intensive aquaculture represent a significant barrier to the advancement of the industry. Advancements in the nutritional modulation of the immune response are anticipated to serve as an effective and cost-efficient alternative for disease management in aquaculture [15]. Examining the impact of feeding rates on fish stress and innate immunity is essential for enhancing our comprehension of fish health and welfare, as well as mitigating growth inhibition and mortality associated with stress and reduced immunity.

Koi (Cyprinus carpio var. koi), a variant of common carp (Cyprinus carpio), is a significant species in the ornamental fish industry due to its rapid growth, ease of cultivation within the cyprinidae family, ornamental attributes, and high market value. Koi is one of the most exportable species. Recent years have seen significant economic losses due to immune suppression and disease outbreaks stemming from improper feeding practices. Furthermore, unlike food fish species, where growth performance and feed efficiency dominate economic returns, koi valuation hinges on esthetic traits (e.g., color vibrancy and fin integrity) and survival longevity. Chronic stress from improper feeding rates may induce morphological abnormalities (e.g., scale loss and body deformities) and immunosuppression, directly reducing its market value. Consequently, feeding regimes for koi demand a distinct evaluation framework that prioritizes health biomarkers over mere growth metrics. We hypothesized that feeding rates deviating from an optimal range would induce chronic stress and suppress innate immunity in koi, with both underfeeding and overfeeding compromising the health status despite divergent mechanisms. This paper aimed to investigate the effects of varying feeding rates on stress and non-specific immunity in koi.

2. Materials and Methods

2.1. Feed

All experimental groups utilized the same feed formulation; however, the feeding rates varied. The feed formulation is presented in

Table 1. Formulation and proximate composition of the experimental diet.

Ingredients (%)
Fish meal	10.00
Soybean meal	24.00
Rapeseed meal	24.00
Wheat gluten	10.30
Wheat flour	26.30
Soybean oil	3.50
Ca(H2PO4)2	1.00
Vitamin premix a	0.15
Mineral premix b	0.15
Choline chloride (50%)	0.20
Lysine (98.5%)	0.30
Methionine	0.10
Proximate composition (% dry-matter basis)
Crude protein	41.27
Crude lipid	6.57
Moisture	6.17
Ash	6.40
Gross energy (MJ/kg)	16.25

Note: ^a Vitamin premix (mg/kg diet): vitamin B₁, l2 mg; vitamin B₂, 5 mg; vitamin B₅, 30 mg; vitamin B₆, 6 mg; vitamin B₁₂, 0.05 mg; vitamin D₃, 5 mg; vitamin E, 40 mg; vitamin K₃, 5 mg; inositol, 100 mg; niacin acid, 35 mg; folic acid, 2 mg, biotin, 0.05 mg; retinol acetate, 25 mg; ascorbic acid, 500 mg; ethoxyquin, 150 mg; corn protein powder, 585 mg. ^b Mineral premix (mg/kg diet): KCl, 10 mg; KI (1%), 3 mg; CoCl₂·6H₂O (1%), 0.35 mg; CuSO₄·5H₂O, 0.7 mg; FeSO₄·H₂O, 20 mg; ZnSO₄·H₂O, 10 mg; MnSO₄·H₂O, 4 mg; Na₂SeO₃·5H₂O (1%), 3.25 mg; MgSO₄·7H₂O, 150 mg; Ca(H₂PO₄)₂·H₂O, 1000 mg; NaCl, 6.8 mg; Zoelite, 292 mg.

and was prepared as extruded feed following the method outline by He et al. (2021) [16]. All ingredients in the formulation were crushed and sieved through an 80-mesh sieve. The powdered feed ingredients were thoroughly mixed in a graded pre-mixing manner. Soybean and fish oils were incorporated, followed by the addition of distilled water added at 15% of the total weight of the feed ingredients. The mixture was transferred to an extruder (model EX1021, Andritz AG, Graz, Austria) at a pelleting temperature of 120 ± 5 °C, resulting in a pellet size of 2.0 mm. The extruded feed was dried at 30 °C and subsequently stored at −20 °C for future use.

2.2. Fish Feeding Trial

The koi utilized in this research were sourced from the Fuyang Quanzhou Ornamental Fish Breeding Base in Anhui, China. Before the start of the experiment, the fish were housed in two indoor recirculation tanks (3.5 × 3.5 × 3.0 m, L:W:H) for one week to acclimate to the culture environment. After the conditioning period, 240 fish (average initial weight: 10.02 ± 0.18 g) were randomly allocated to 24 net cages (1.0 × 1.0 × 1.0 m, L:W:H) suspended in indoor recirculation tanks (3.5 × 3.5 × 3.0 m, L:W:H), with 10 fish per net cage. The 24 net cages were randomly assigned to 6 groups, with 4 parallels per group. Each group received feed corresponding to 1, 2, 3, 4, 5, and 6% of their body weights, respectively. The feeding rate gradient (1–6% BW/day) was established based on three key considerations: (1) Previous studies suggest that optimal feeding rates for cyprinids typically range between 2 and 5% BW/day [10,17]; thus, extending the gradient beyond this range allows for the detection of underfeeding and overfeeding effects. (2) Previous studies on feeding rates in cyprinid typically ranged from 2 to 7% BW/day, but all studies used pelleted feed. Our previous studies found that extruded feed significantly reduced fish feed intake compared to pelleted feed under the same formula [16,18]. This study used extruded feed, so the upper limit of the feeding rate was lowered from 7% BW/day to 6 BW/day. (3) The wide span accommodates potential size-specific metabolic demands during the rapid growth phase. Fish in each cage were weighed biweekly, and the feed quantity was modified based on the most recent weight. The fish received feed three times daily at 08:00, 12:00, and 16:00, with a culture cycle lasting eight weeks. Dissolved oxygen, water temperature, pH, and total ammonia nitrogen during the culture period were measured at 25 ± 2 °C, 5.0–6.0 mg/L, 7.0–7.4, and 0.04 mg/L, respectively.

2.3. Sample Collection Methodology

At the conclusion of the culture period, fish were subjected to a 24 h starvation period. The fish in each cage were weighed to calculate growth data. Following this, the fish were thoroughly anesthetized using eugenol (St. Louis, MO, USA, >98% purity) at a concentration of 0.1 g per liter of water. A single fish was randomly chosen from each cage, and blood was collected from the tail vein using a 1 mL sterile syringe, which was subsequently injected into a 1.5 mL anticoagulation tube. The anticoagulated tube with blood was positioned in a centrifuge pre-cooled to 4 °C and centrifuged at 3000 rpm for 10 min. The supernatant was aspirated to assess stress, liver injury, and immune-related biochemical indicators in the plasma. Following blood sampling, the fish were dissected on an ice pack, and the liver and spleen were promptly separated. The liver was then stored in two portions at −20 °C and −80 °C for subsequent analysis of antioxidant-related biochemical and molecular indices, respectively. The spleen was preserved at −80 °C for the assessment of inflammation and immunity-related molecular markers.

2.4. Indicators of Stress and Liver Injury

Plasma samples were removed from the −20 °C refrigerator and gradually thawed in a 4 °C environment for preservation purposes. Levels of cortisol (COR, Cat No. H094-1-1), glucose (GLU, Cat No. F006-1-1), and lactate (LAC, Cat No. A019-2), as well as activities of alanine aminotransferase (ALT, Cat No. C009-2-1) and aspartate aminotransferase (AST, Cat No. C010-2-1), were measured using commercial kits from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). All procedures were executed in strict compliance with the manufacturer’s guidelines.

2.5. Indicators of Antioxidant Activity

Liver samples were removed from the −20 °C refrigerator and thawed gradually in a 4 °C environment. The thawed liver samples were homogenized with nine times the volume of physiological saline (w/v, 1:9) at 4 °C. The homogenate underwent centrifugation at 4 °C for 10 min at 3000 rpm, and the supernatant was collected for the assessment of liver antioxidant-related indicators. The activities of total superoxide dismutase (t-SOD), glutathione peroxidase (GPX), and catalase (CAT) were assessed using the methodology outlined by Lygren et al. (1999) [19]. The content of malonaldehyde (MDA) was assessed using the thiobarbituric acid method.

2.6. Indicators of Immunity

Plasma samples were removed from the −20 °C refrigerator and gradually thawed at 4 °C for preservation purposes. The plasma levels of immunoglobulin M (Ig M), complement 3 (C3), and complement 4 (C4) were assessed using the enzyme-linked immunosorbent assay [20,21]. The activities of plasma lysozyme (LZM), myeloperoxidase (MPO), and acid phosphatase (ACP) were assessed using the methodologies outlined by Obach et al. (1993) [22], Siwicki et al. (1994) [23], and Lulijwa et al. (2019) [24], respectively. Total protein (TP) and albumin (ALB) levels in plasma were measured using commercial kits based on the Biuret method and the BCG dye-blinding method (Qualigens Diagnostics, Mumbai, India), respectively. The globulin (GLO) level was calculated by subtracting the ALB level from the TP level.

2.7. Antioxidant and Immune-Related Gene Expression

Tissue selection: Previous studies indicate that the contents of tnf-α, tgf-β, il1, il10, hsp70, igm, and igt1 are abundant in the spleen [25,26], while SOD and CAT are predominantly found in the liver [27]. Consequently, the spleen and liver were chosen for total RNA extraction.

RNA extraction involved the collection of 100 mg of spleen and liver tissues, which were placed into separate centrifuge tubes containing 1 mL of Trizol (Invitrogen, Carlsbad, CA, USA). Total RNA extraction was performed utilizing the RNAios Plus kit (Takara, Shiga, Japan). RNA quality was subsequently evaluated utilizing a bioanalyzer (Agilent, Santa Clara, CA, USA, model 2100). RNA was then diluted to a uniform concentration for reverse transcription.

RNA reverse transcription involved processing RNA using a reverse transcription kit (Takara, Shiga, Japan) to synthesize cDNA. The reaction was conducted in two cycles: 40 min at 42 °C and 2 min at 90 °C. The samples were subsequently stored at 4 °C.

RT-qPCR was performed using 2 × ChamQ universal SYBR qPCR master mix (Vazyme, Nanjing, China). The reaction protocol comprised 40 cycles: 30 s at 95 °C for denaturation, 5 s at 95 °C for extension, and 40 s at 60 °C for annealing. Gene-specific amplification was confirmed through solubility profile analysis.

The design of primers and the assessment of relative expression: Primers for tnf-α, tgf-β, il1, il10, hsp70, igm, igt1, SOD, and CAT were created using Primer Premier 5.0, with β-actin serving as the housekeeping gene [25,26,27]. Primers were synthesized by Shanghai Generay Biotech Co., Ltd. (Shanghai, China), as listed in

Table 2. Nucleotide sequences for real-time PCR primers.

Target Gene	Forward (5′-3′)	Reverse (5′-3′)	References
tnf-α	GTGATGGTGTCGAGGAGGAAG	TCTGAGACTTGTTGAGCGTGAA	[27]
tgf-β	CCTGGGCTGGAAGTGGATAC	GTAAAAGATGGGCAGTGGGTC	[25]
il1	GATGCAAATGCCCTCAAATACA	GGCTCTTGACGTTCCTTTTG	[25]
il10	GGAGGGCTTTCCAGTGAGAC	TGTTGCACGTTTTCGTCCAG	[25]
hsp70	GTGTCCATCCTGACCATTGAAGA	CTGACTGATGTCCTTCTTGTGCTTC	[25]
igm	CACAAGGCGGGAAATGAAGA	GGAGGCACTATATCAACAGCA	[26]
igt1	AAAGTGAAGGATGAAAGTGT	TGGTAACAGTGGGCTTATT	[26]
sod	GATGGCAGCCTTGGAAGTGAC	TCAGAACAATCAGGAAGGAGGAA	[27]
cat	CTGGAAGTGGAATCCGTTTG	CGACCTCAGCGAAATAGTTG	[27]
β-actin	GCTATGTGGCTCTTGACTTCGA	CCGTCAGGCAGCTCATAGCT	[27]

Note: tnf-α, tumor necrosis factor α; tgf-β, transforming growth factor β; il1, interleukin 1; il10, interleukin 10; hsp70, heat shock protein 70; igm, immunoglobulin M; igt1, immunoglobulin t1; sod, superoxide dismutase; cat, catalase.

, with only those exhibiting amplification efficiencies exceeding 90% being utilized. The relative expression of all genes was normalized using the housekeeping gene β-actin, and relative gene expression was evaluated using the 2^−∆∆CT method [28].

2.8. Calculation and Statistical Analysis

In this study, the growth parameter was computed as follows:

Daily gain index (DGI, %/day) = (final body weight^1/3 − initial body weight^1/3) × 100/days fed [29].

The data passed normality and homogeneity tests. Data were analyzed via one-way ANOVA using the SPSS (25.0) program. When data were significantly different (p < 0.05), means were ranked using Tukey’s HSD multiple range test. Data in the article are expressed as the mean ± standard error of four repetitions.

3. Results

3.1. Growth

No mortality was recorded in the experimental fish during the entire culture period. DGI exhibited a notable increase, with the feeding rate rising from 1% to 3% BW/day (p < 0.05). No significant difference in DGI was observed between the 3% and 4% BW/day groups (p > 0.05); however, DGI decreased significantly with increasing feeding rates (p < 0.05) (Figure 1).

3.2. Stress and Liver Injury Indicators

Figure 2 illustrates that varying feeding rates did not significantly influence the levels of LAC or the activities of AST and ALT (p > 0.05). The COR levels exhibited an initial decline followed by an increase as the feeding rate rose from 1% to 6% BW/day, with the 4% BW/day group demonstrating a significantly lower level compared to the other groups (p < 0.05). The GLU levels exhibited a significant increase (p < 0.05), with an increase in the feeding rate from 1% to 3% BW/day, which subsequently stabilized (p > 0.05).

3.3. Antioxidant Indicators

Figure 3 illustrates that as the feeding rate increased from 1% to 6% BW/day, GPX activity initially rose and subsequently declined, while MDA content first decreased and then increased; however, no significant differences were noted (p > 0.05). The activities of t-SOD and CAT exhibited a significant increase (p < 0.05) as the feeding rate rose from 1% to 3% BW/day. The increase in feeding rate from 3% to 6% BW/day resulted in a significant decrease in t-SOD activity (p < 0.05), whereas CAT activity tended to stabilize (p > 0.05).

Figure 1. Daily gain index of koi carp fed at various feeding rates. Bars assigned with different superscripts are significantly different (p < 0.05).

Figure 1. Daily gain index of koi carp fed at various feeding rates. Bars assigned with different superscripts are significantly different (p < 0.05).

3.4. Immunity Indicators

Table 3. Plasma haemato-immunological parameters of koi carp fed at various feeding rates.

Feeding Rates (% BW/Day)	1	2	3	4	5	6
LZM (U/mL)	87.53 ± 3.23 c	104.07 ± 3.78 bc	112.01 ± 3.71 bc	114.86 ± 1.77 ab	117.44 ± 9.24 ab	139.69 ± 9.06 a
C3 (mg/L)	260.35 ± 13.45 b	315.64 ± 27.58 ab	339.19 ± 14.18 ab	361.65 ± 29.87 ab	378.43 ± 23.24 a	377.08 ± 26.54 a
C4 (mg/L)	190.36 ± 10.53 b	226.69 ± 19.00 ab	273.24 ± 20.10 a	277.10 ± 20.16 a	283.43 ± 10.68 a	291.62 ± 15.71 a
IgM (g/L)	1.89 ± 0.11 a	1.97 ± 0.05 a	1.99 ± 0.03 a	1.97 ± 0.02 a	2.01 ± 0.03 a	2.12 ± 0.02 a
MPO (U/L)	7.41 ± 0.27 b	8.36 ± 0.30 b	14.85 ± 0.64 a	17.51 ± 1.04 a	14.92 ± 1.17 a	15.48 ± 0.54 a
ACP (U/L)	194.63 ± 5.78 a	191.35 ± 6.56 a	212.01 ± 11.19 a	210.31 ± 5.11 a	187.98 ± 8.91 a	213.68 ± 6.20 a
TP (g/L)	16.32 ± 0.65 b	31.90 ± 0.86 a	32.75 ± 1.21 a	34.41 ± 1.85 a	34.51 ± 1.45 a	32.96 ± 1.09 a
ALB (g/L)	11.51 ± 0.38 b	17.61 ± 1.37 a	17.75 ± 1.27 a	17.43 ± 1.10 a	17.81 ± 1.30 a	16.88 ± 0.79 a
GLO (g/L)	4.81 ± 0.77 b	14.30 ± 2.07 a	14.99 ± 0.73 a	16.98 ± 1.70 a	16.70 ± 2.57 a	16.09 ± 0.83 a

Note: LZM, lysozyme; C3, component 3; C4, component 4; IgM, immunoglobulin M; MPO, myeloperoxidase; ACP, phosphatase; TP, total protein; ALB, albumin; GLO, globulin. Significant differences (p < 0.05) among feeding rates within each treatment are indicated by different letters.

indicates that IgM content and ACP activity remained unaffected by the feeding rate (p > 0.05). LZM activity and C3 content exhibited a significant increasing trend as the feeding rate rose from 1% to 4% BW/day (p < 0.05), but no significant differences were observed with further increases in the feeding rate (p > 0.05). The C4 content and MPO activity exhibited a significant increase (p < 0.05) as the feeding rate rose from 1% to 3% BW/day. None of the three indicators demonstrated a statistically significant difference (p > 0.05) as the feeding rate further increased. The contents of TP, ALB, and GLO exhibited a significant upward trend as the feeding rate increased from 1% to 2% BW/day, which subsequently stabilized (p > 0.05).

Figure 2. Plasma stress response and diagnosis of liver injury in koi carp fed at various feeding rates. Bars assigned with different superscripts are significantly different (p < 0.05). (A), cortisol (COR); (B), glucose (GLU); (C), lactate (LAC); (D), aspartate aminotransferase (AST); (E), alanine aminotransferase (ALT).

Figure 2. Plasma stress response and diagnosis of liver injury in koi carp fed at various feeding rates. Bars assigned with different superscripts are significantly different (p < 0.05). (A), cortisol (COR); (B), glucose (GLU); (C), lactate (LAC); (D), aspartate aminotransferase (AST); (E), alanine aminotransferase (ALT).

Figure 3. Hepatic oxidative status of koi carp fed at various feeding rates. Bars assigned with different superscripts are significantly different (p < 0.05). (A), total superoxide dismutase (t-SOD); (B), catalase (CAT); (C), glutathione peroxidase (GPX); (D), malonaldehyde (MDA).

Figure 3. Hepatic oxidative status of koi carp fed at various feeding rates. Bars assigned with different superscripts are significantly different (p < 0.05). (A), total superoxide dismutase (t-SOD); (B), catalase (CAT); (C), glutathione peroxidase (GPX); (D), malonaldehyde (MDA).

3.5. Antioxidant and Immune-Related Gene Expression

Figure 4 illustrates that the feeding rate had no significant impact on the expression levels of tgf-β and hsp70 (p > 0.05). As the feed rate increased from 1% to 6% BW/day, the expression levels of tnf-α, il1, and il10 exhibited a significant trend of initially decreasing and subsequently increasing (p < 0.05), with minimum values recorded at 4%, 3%, and 5% BW/day, respectively. As the feed rate increased from 1% to 6% BW/day, the expression levels of igm and igt1 exhibited a continuous upward trend. Significant differences were observed at feed rates of 5% and 4% BW/day (p < 0.05), respectively, after which the levels tended to stabilize (p > 0.05). Increasing the feeding rate from 1% to 4% BW/day resulted in a significant increase in the expression levels of SOD and CAT (p < 0.05). However, further increasing the feeding rate from 4% to 6% BW/day led to a significant decrease (p < 0.05).

4. Discussion

In artificial breeding environments, artificially formulated feed serves as the exclusive source of nutrients necessary for fish growth. The feeding rate significantly influences the growth performance of fish [30,31]. This study demonstrates that the feeding rate significantly influences the growth of koi. The DGI of the fish reached its peak at a feeding rate of 3–4% BW/day. The weight gain of koi fed at 1–2% BW/day and 5–6% BW/day was significantly lower compared to that of koi fed at 3–4% BW/day. This suggests that a feeding rate of 3–4% body weight per day is optimal for growth performance in this species, while both low and high feeding rates can negatively impact growth performance in koi. This finding is consistent with trends observed in other cyprinids [10,17]. From a long-term perspective, sustained suboptimal feeding may exacerbate these effects. Chronic underfeeding could lead to irreversible growth stunting due to developmental plasticity constraints [32], while prolonged overfeeding might induce metabolic disorders such as hepatic steatosis, as observed in chronically overfed zebrafish [33]. Three factors account for growth retardation under low feeding conditions: (1) Inadequate feed intake fails to satisfy the nutritional needs of koi, leading to reduced growth performance [34]. (2) Underfeeding causes koi to congregate at the surface of the water and compete for expanded feed, a behavioral pattern that causes physiological stress, which in turn inhibits growth [35]. (3) Enhanced swimming activity during feed competition leads to increased energy expenditure. Furthermore, growth retardation under elevated feeding rates can be attributed to two primary factors: Overfeeding imposes a significant strain on the intestines of fish. The intervention not only decreases the intestinal fold area [36] but also induces oxidative stress in the intestine, compromising the integrity of the intestinal mucosa, particularly in fish lacking a stomach, such as koi [37]. Persistent intestinal oxidative damage may predispose fish to chronic enteritis and impaired nutrient absorption over time [38], potentially reducing the lifespan in ornamental specimens. Additionally, overfeeding disrupts the activity of the GH-IGF-1 axis, thereby inhibiting fish growth [10]. The 3–4% BW/day feeding rate maximizes koi weight gain while avoiding intestinal damage from overfeeding.

This study found no effect of the feeding rate on LAC content, AST, and ALT activity. This suggests that the feeding rate has no impact on metabolic efficiency and liver health in koi. Fish typically utilize anaerobic metabolism to fulfill energy demands in unfavorable conditions. LAC is the final product of anaerobic glycolysis, and increased levels of lactate typically signify reduced metabolic efficiency [39]. AST and ALT serve as specific indicators of liver damage, as they are typically released into the bloodstream when the function and integrity of hepatocytes are compromised. Increased serum activities of AST and ALT typically signify liver damage [40]. This study found that the COR content was highest in the group receiving a feeding rate of 1% BW/day, whereas the lowest GLU level was also observed in this group. This contradiction can be elucidated by the following reasons. Insufficient feed leads to increased competition among fish for food, resulting in heightened stress levels [8,10]. COR serves as a sensitive indicator for evaluating the stress state of animals [41], and an increase in its levels supports this perspective. Moreover, under typical conditions, fish stress correlates with elevated GLU levels [42], whereas the COR content in the 1% BW/day group, which exhibited the highest COR levels in this study, was the lowest. The GLU content of fish is influenced by stress, carbohydrate intake, and metabolism at varying feeding rates [8]. At low feeding rates, fish exhibit reduced food intake and inadequate carbohydrate consumption. Under stress conditions, fish demonstrate increased energy demands and quickly utilize glucose for energy [43]. Consequently, fish blood glucose may decrease under stress conditions [44]. An adequate food supply (4% BW/day) helps to alleviate the stress caused by koi competing for feed and maintains a steady supply of energy. However, to further clarify the inconsistency in the glucose–cortisol dynamics observed in the 1% BW/day group, a larger sample size needs to be included in future work.

Previous studies confirm that alterations in nutritional status and stress induced by feed can influence the antioxidant capacity of fish [45,46]. Furthermore, antioxidant status is closely linked to the physiological function and immune capacity of fish [47]. It is essential to examine alterations in antioxidant capacity when evaluating the impact of the feeding rate on the health status of koi. The current study demonstrated that t-SOD and CAT activities exhibited a significant increase as the feeding rate rose from 1% to 3% BW/day, followed by a decline with additional increases in the feeding rate. The MDA content exhibited an inverse relationship with t-SOD and CAT activities; however, no significant difference was observed. This suggests that both low and high feeding rates negatively affect the antioxidant capacity of koi, yet neither leads to significant oxidative damage. However, the cumulative effects of prolonged antioxidant system overload warrant consideration. The chronic suppression of SOD-CAT activity at extreme feeding rates may accelerate cellular senescence through telomere shortening mechanisms [48]. Under standard physiological conditions, the generation and elimination of reactive oxygen species (ROS) in fish maintain a dynamic equilibrium. Under stress conditions, fish produce significant quantities of ROS. If not promptly neutralized by the antioxidant system, these ROS can lead to the damage of lipids, proteins, and DNA, resulting in issues such as lipid peroxidation and damage to proteins and DNA [49]. In response to these adverse conditions, fish activate antioxidant mechanisms to mitigate ROS overload and consequently diminish the detrimental effects on cells and tissues [50]. The SOD-CAT system serves as the primary defense mechanism against oxidative stress, as it functions to scavenge ROS [51]. MDA serves as a biomarker for oxidative damage [52]. Previous studies indicate that feeding levels significantly influence the degree and duration of increased oxygen consumption post-meal in fish [11,53]. This phenomenon affects the production of ROS and subsequently the activity of SOD and CAT. Feeding can be a persistent long-term stressor for fish [11,54], and the range of feeding rates established in this study was limited. No notable differences in oxidative damage were detected among koi subjected to varying feeding rates. Nevertheless, the observed inflammatory gene upregulation (tnf-α, il1) under both feeding extremes suggests potential for chronic low-grade inflammation. The 3–4% BW/day feeding rate improved the scavenging of reactive oxygen species by koi and prevented the development of chronic low-grade inflammation.

Cyprinidae fish are economically viable and are bred to supply consumers with high-quality protein. Their feeding programs emphasize feed conversion and growth rates. Koi, ornamental fish derived from common carp, are costly and cultivated for decorative purposes; consumers express concerns regarding their health and longevity. Consequently, it is imperative to focus on feeding programs that enhance the immune capabilities of this species. Research indicates that feeding influences the development of various metabolic traits and immune functions in fish [55]. This study indicates that the lowest feeding rate led to reduced haemato-immunological parameters in comparison to moderate feeding rates. This suggests that feeding restriction may cause immunosuppression in koi, as these humoral and cellular factors are critical components of the fish’s non-specific immune system and are essential for pathogen defense and overall health [56]. Long-term immunosuppression could fundamentally alter disease resistance dynamics in ornamental populations, potentially increasing susceptibility to endemic pathogens like Cyprinid herpesvirus. The observed immunosuppression may result from inadequate energy intake by the fish due to low feeding rates. Cellular and humoral immunity represent physiological responses that require significant energy expenditure [13]. The immune system’s proper functioning necessitates adequate energy as well as proteins, amino acids, fats, essential fatty acids, carbohydrates, vitamins, minerals, and certain non-nutrient dietary compounds [15]. Insufficient feeding rates can lead to a deficiency of these substances, adversely affecting the immune system. Moreover, the fish receiving the highest level of feeding did not exhibit a significant enhancement in haemato-immunological parameters when compared to those that were moderately fed, suggesting that overfeeding does not confer advantages to the immune defense of this species. It has been proposed that supplying additional nutrients beyond the optimal level for fish growth may improve immune function [13,15,57]. This contradiction can be elucidated by two factors: The aforementioned perspective is not universally applicable to all species. The immune system of chinook salmon (Oncorhynchus tshawytscha) exhibits significant adaptability to substantial fluctuations in metabolic energy. The phagocytic function of cellular immunity demonstrates an inverse relationship with the amount of feed [5]. Furthermore, the studies supporting this perspective primarily focused on the addition or enhancement of specific nutrients, whereas this study increased the feeding rate. The additional intake of various nutrients by fish using the same feed formula increased in equal proportions, contrasting with studies that focus on the individual enhancement of one or several nutrients.

We assessed the expression of immune and antioxidant-related genes to elucidate the molecular mechanism underlying the impact of feeding rate on the health status of koi. This study examines the expression of tnf-α, il1, and il10 in both low-feeding and high-feeding groups. This suggests that both low and high feeding rates result in an elevated inflammatory response and a diminished innate immune capacity in koi. tnf-α is an inflammatory factor produced by macrophages/monocytes [58] and acts as a potent proinflammatory cytokine [59] that regulates the inflammatory response by limiting its extent and duration [60]. il1 is an inflammatory regulator that activates various cellular innate immune responses. Similar to tnf-α, it can initiate systemic inflammation and serves as a crucial element of the innate immune response [61]. il10 is a very effective anti-inflammatory and immunosuppressive cytokine [62], and it plays a dual role in supporting humoral immunity and inhibiting inflammatory responses [63]. It also functions to inhibit the synthesis of IL-1 and TNF-α [64]. The expression levels of sod and cat exhibited a trend opposite to that of tnf-α, il1, and il10. Both low and high feeding rates negatively affect the antioxidant capacity of koi. SOD and CAT serve as the primary defense mechanisms against oxidative damage from ROS and play a direct role in their neutralization [65]. Under varying degrees of oxidative stress, the expression levels of sod and cat exhibited two distinct trends: (1) Within a specific range, the expression levels of sod and cat were upregulated in response to increasing ROS levels to mitigate oxidative stress and inhibit further ROS accumulation [62]. (2) An excessive buildup of ROS was associated with the activation of the molecular antioxidant system, resulting in the downregulation of SOD and CAT gene expression [66]. The findings of this study, in conjunction with prior physiological and biochemical indicators, indicate that the trend of sod and cat expression aligns with the physiological state outlined in (1). Additionally, the expression levels of igm and igt1 exhibited an increasing trend as the feeding rate rose from 1% to 6% BW/day; however, no significant difference was observed in the expression levels of igm and igt1 between fish fed at the highest rate and those fed at a moderate rate. Low feeding rates negatively affect the immune system of koi, while high feeding rates do not enhance immunity. Supporting evidence: (1) IgM serves as a primary element of the humoral immune system in bony fish, fulfilling essential defense roles, whereas Igt1 primarily contributes to mucosal immunity. The expression levels of igm and igt1 can partially indicate the functionality of adaptive and mucosal immunity [26]. (2) Additionally, the energy necessary for immune system development is contingent upon the body’s energy budget [14]. Insufficient energy intake adversely affects the immune system. (3) Conversely, when energy intake surpasses the requirements for immune system development and function, excess energy may be utilized to augment nutrient reserves, thereby mitigating potential starvation risks [67]. Maintaining an appropriate feeding amount (not less than 3% BW/day) is essential for maintaining innate immunity and helps koi resist common pathogens.

Finally, the experimental design used four replicates per treatment (n = 4 cages, with 10 fish per cage), which aligns with standard aquaculture controlled trial protocols [10,11,16]. Nevertheless, we recognize that increasing the number of replicates to n = 8 could enhance the statistical power of the observed metrics.

5. Conclusions

This study demonstrates that the feeding rate significantly influences the growth rate, stress response, antioxidant levels, and non-specific immune capacity of koi. Reduced feeding rates (1–2% BW/day) result in stunted growth and diminished antioxidant and immune capacity in this species. Overfeeding (5–6%) does not enhance the immune defense of koi and may result in diminished growth performance, reduced antioxidant capacity, and heightened inflammatory response. These findings provide a reference for the scientific feeding of koi and make it clear that feeding intensity is not linearly related to the growth rate and health. In addition to body size (which is related to the growth rate), the vividness of body color is also a major selling point for koi, and further research could be carried out to investigate the effects of feeding rates on carotenoid deposition and changes in body coloration in koi.