Evaluation of Seaweed Meal and Konjac Glucomannan Mixture as Feed Ingredients in Largemouth Bass Micropterus salmoides

Abstract

To address the negative effects of high-starch diets on largemouth bass (LMB), this study evaluated the feasibility of using a Gracilaria lemaneiformis-konjac glucomannan mixture (GKM, 2:1) as a substitute for strong flour (SF). Four iso-nitrogenous and iso-lipid diets were formulated: a control (15% SF; GK00) and three other diets replacing 33.3% (GK05), 66.7% (GK10), or 100% (GK15) of SF with GKM. Each diet was randomly administered to triplicate tanks of fish (10.49 ± 0.232 g) for a 10-week feeding trial. Results showed that the GKM inclusion groups significantly improved the fish survival and feed intake. Fish in GK05 and GK10 groups exhibited significantly higher final body weight, weight gain, and specific growth rate than the GK00 group, while GK15 showed no significant increase in these metrics. There was no impairment in protein, lipid, phosphorus, and energy retention efficiency in the GK05 and GK10 groups compared to those of the GK00 group. Apparent digestibility for feed dry matter, protein, lipid, phosphorus, and the 16 amino acids was not decreased in the GK05 and GK10 groups relative to the GK00 group. In addition, this study revealed reduced phosphorus waste per kilogram of weight gain in GK05 and GK10. In conclusion, these findings position GKM as a sustainable alternative to SF in feed for LMB.

1. Introduction

Starch is a primary component frequently used in contemporary aquatic animal feed, serving as the key binder in aquatic feed [1]. As a cost-effective energy source, it has been demonstrated that an appropriate starch supplementation in fish feed could reduce protein consumption for energy supply, thereby achieving a protein-sparing effect [2]. The reduction in protein degradation for energy acquisition resulting from the appropriate content of starch supplementation could enhance the protein synthesis metabolism of fish and consequently decrease nitrogen emissions into the environment [3]. By potentially sparing protein for growth and other vital functions, they offer a more economical alternative to proteins and lipids, making them an attractive dietary component in aquaculture. However, many studies have shown that long-term consumption of high-starch feed would impair fish growth and feed utilization efficiency, especially for carnivorous fish, which commonly have a lower ability to utilize dietary starch compared to herbivorous or omnivorous fish [4].

Largemouth bass (LMB), Micropterus salmoides, a quintessential carnivorous fish, is one of the most favored freshwater aquaculture species due to its significant economic value and demand in recreational fishing. Over recent years, its farming production has witnessed a steady and continuous increase, with outputs rising from 170,000 tons in 2009 to 802,486 tons by 2022 in China [5]. Previous studies have reported protein requirements of 45–50% and lipid requirements of 9–16% for LMB [6,7,8]. Regarding carbohydrates, while specific requirements remain debated, consensus exists that long-term consumption of high-starch feed led to a decline in the growth traits of LMB. Consequently, research recommended that the starch content in their feed should be reduced to 10% [9,10,11,12]. However, despite clear recommendations to limit dietary starch in carnivorous fish, practical feed formulations often exceed optimal levels due to processing requirements and economic constraints. Thus, identifying functional starch substitutes that maintain feed structure, enhance nutrient utilization, and reduce environmental waste is urgently needed [13,14,15,16].

Macroalgae, also called seaweeds, contain not only protein but also other functional components such as functional non-starch polysaccharides (NSPs), pigments, carotenoids, phenolics, fatty acids, vitamins, and other secondary metabolites [17,18]. Despite the incomplete characterization of their bioactive components, macroalgae are increasingly employed as sustainable functional ingredients or feed additives in aquafeeds, leveraging their unique polysaccharide profiles and micronutrients to enhance fish growth performance while mitigating aquacultural nutrient discharge [14,19,20]. Gracilaria lemaneiformis (GL) is a commercial-scale farming red alga in China; its cultivation yield has increased from 398,920 tons in 2021 to 610,824 tons in 2022, according to the China Fisheries Yearbook 2023 [5]. The massive cultivation yield means that there was a significant amount of non-edible GL or processing byproducts generated each year, and this promises them the potential to be used in formula feed. Konjac glucomannan (KG) has been used as the feed binder in fish feed processing for its characteristics of high cohesiveness, high suspension, and high water-holding properties, with non-toxic and harmless properties to farmed animals [21]. Furthermore, positive functions of KG added to diets for carnivorous fish have also been found [22]. However, complete replacement of starch with GL is difficult to meet the processing characteristics of fish feed, while the price of KG is relatively high. Therefore, a hypothesis that the strategic incorporation of GL and KG in aquafeed formulations synergistically replicates starch’s functional properties while delivering prebiotic benefits and maintaining cost efficiency was formulated. To validate this hypothesis, a 10-week feeding trial was conducted to assess the effects of varying dietary levels of the GL meal and KG mixture by replacing the starch in this study.

2. Materials and Methods

2.1. Feed Formulation and Diet Preparation

To investigate the responses of LMB to the graded levels of GKM (a 2:1 mixture of GL and KG), four iso-nitrogenous (49.46 ± 0.763%) and iso-lipid diets (10.75 ± 0.421%) were formulated, including a control diet (GK00) with 15% strong flour (SF, from wheat source) and other three diets (GK05, GK10, and GK15) with varying levels of SF replaced by GKM at the ratio of 33.3%, 66.7% and 100%, respectively. The proximate composition of the main feed ingredients, including the contents of moisture, crude protein (Kjeldahl method), crude lipid, ash, and crude fiber, was measured following a series of Chinese standards (GB/T 6435-2014, GB/T 6432-2018, GB/T 6433-2006, GB/T 6438-2007, and GB/T 6434-2022) [23,24,25,26,27]. The results are detailed in

Table 1. Proximate composition of the main ingredients used in this experiment (%).

Items	FM 1	SBM 2	SPI 3	SF 4	GL 5	KG 6
Moisture	9.32	11.29	6.79	12.61	8.53	7.26
Crude protein	65.27	44.24	84.76	15.83	18.40	3.98
Crude lipid	8.01	0.28	2.64	1.43	2.20	0.20
Ash	17.32	6.74	4.78	1.21	19.82	1.49
Crude fiber	/	5.90	0.20	0.90	6.10	1.70
NFE 7	0.08	31.55	0.83	68.02	44.95	85.37

¹ Fish meal. Pesquera Exalmar S.A.A. ² Soybean meal. COFCO Dongguan Grain and Oil Industry Co., Ltd., Dongguan, Guangdong, China. ³ Soy protein isolates. Shansong Biological Co., Ltd., Linyi, Shandong, China. ⁴ Strong flour. Wudeli Xinghua Flour Factory, Xinghua, Jiangsu, China. Starch content in SF was calculated as the formula that starch = 100 − (Moisture% + Crude Protein% + Crude Lipid% + ash% + crude fiber%)]. ^5, Gracilaria lemaneiformis meal; Ningxia Vanilla Biotechnology Co., Ltd., Guyuan, Ningxia, China. ⁶ Konjac glucomannan mannan. Ningxia Vanilla Biotechnology Co., Ltd., Guyuan, Ningxia, China. ⁷ Nitrogen free extract calculated according to the formula that NFE = 100 − (Moisture% + Crude protein% + Crude fat% + Ash% + Crude fiber%).

.

For diet preparation, large-size feed ingredients were crushed through a 100-mesh sieve and fully mixed according to the feed formula. Subsequently, water was blended into the mixture using a mixing device (ND-VH-2, Jingjiang Huiheng Machinery Manufacturing Co., Ltd., Wuxi, China). The moist mash obtained was pre-gelatinized at 100 °C for 30 s before pelleting through a 3 mm die using a two-section twin-screw extruder (DSE65, Jinan Dingrun Technology Development Co., Ltd., Jinnan, China). The produced feed pellets were dried in a forced-air-drying oven (45 °C) until the moisture content was below 8%, following fish oil coating in a revolving drum (DX-800, Jinan Dingrun Technology Development Co., Ltd., Jinnan, China). The diets were stored at −20 °C until fish feeding. The measurement of dietary moisture, crude protein, crude lipid, and ash contents was similar to that used for the feed ingredients, and dietary phosphorus was measured according to Chinese Standard GB/T 6437-2018 [28]. The feed formula and diet proximate composition are presented in

Table 2. Feed formulation and proximate composition of experimental diets.

Items	GK00	GK05	GK10	GK15
Ingredients
Fish meal (g/kg)	480.0	480.0	480.0	480.0
Soybean meal (g/kg)	170.0	170.0	170.0	170.0
Soy protein isolates (g/kg)	45.0	50.0	50.0	50.0
Fish oil (g/kg) 1	75.0	75.0	75.0	75.0
Premix (g/kg) 2	20.0	20.0	20.0	20.0
Monocalcium phosphate (g/kg) 3	10.0	10.0	10.0	10.0
Choline chloride (g/kg) 4	3.0	3.0	3.0	3.0
Yttrium (III) oxide (g/kg) 5	3.0	3.0	3.0	3.0
Strong flour (g/kg)	150	100.0	50.0	0.00
GL (g/kg)	0.0	34.0	67.0	100.0
KG (g/kg)	0.0	16.0	33.0	50.0
Medical stone (g/kg) 6	44.0	39.0	39.0	39.0
Composition
Dry matter (g/kg)	935.2	931.7	930.5	922.5
In dry matter
Crude protein (g/kg)	493.2	503.5	496.4	485.1
Crude lipid (g/kg)	113.6	104.2	106.5	105.5
Ash (g/kg)	171.5	179.2	180.1	186.8
Phosphorus (g/kg)	18.2	17.4	16.6	16.2
Starch (g/kg) 7	116.8	77.8	38.92	0
Gross energy (MJ/kg) 8	18.14	17.34	16.57	15.58

¹ Rongcheng Haida Fish Meal Co., Ltd., Weihai, Shandong, China. ² Multidimensional and multimineral premix for largemouth bass, provided by Guangdong Bide Biotechnology Co., Ltd., Guangzhou, Guangdong, China. ³ Feed grade, Guizhou Yuedu Chemical Co., Ltd., Qiannan Buyi and Miao Autonomous Prefecture, Guizhou, China. ⁴ NB Group Co., Ltd., Zouping, Shandong, China. ⁵ 99.99% metals basis, Weng Jiang Regent Co., Ltd., Guangzhou, Guangdong, China. ⁶ Kexu Jiancai Co., Ltd., Shijiazhuang, Hebei, China. ⁷ Calculated as follows: Starch (g/kg) = strong flour content in feed formula (g/kg) × starch content in strong flour (%)/100. ⁸ Calculated values according to Zhang et al. [12], Gross energy (kJ/kg) = protein content (g/kg) × 23.9 kJ/g + lipid content (g/kg) × 37.8 kJ/g + starch content (g/kg) × 17.6 kJ/g.

.

2.2. Fish Management and Feeding

The fish used in this study were purchased from the Guangzhou Zhongyu Fishery Seeding Co., Ltd. (Guangzhou, Guangdong, China). After 1 month of acclimation, a 10-week feeding trial was conducted with an indoor aquaculture system. Before the feeding trial, the juvenile LMB was fasted for 24 h first, after which 432 healthy fish of initial body weight (IBW) of (10.49 ± 0.232) g were hand-stored into 12 cleaned fiberglass cylindrical tanks (diameter × height = 100 cm × 75 cm, water vol. about 470 L) with 36 fish per tank. The applied management protocol in the feeding trial was according to Cheng et al. [29] with slight adjustments. In detail, throughout the entire feeding trial, the fish management strategies, apart from the feeding diets, remained consistent. In the first 2 weeks, 10% of water was exchanged daily in each tank using dechlorinated tap water. In the following days, the daily water exchange amount was increased by 10% every 2 weeks until the maximum amount reached 55%. An artificial photoperiod (12 h light, 12 h dark) was implemented, and the average water temperature was 30.4 ± 2.4 °C (range from 27.4 °C to 36.0 °C) throughout the feeding trial.

During the feeding trial, each feed was provided to three tanks of fish two times per day (08:30 and 19:00) by hand to apparent satiety. Before each feeding, the water inlet and aeration were stopped to facilitate the observation of fish feeding, and each feeding was set for 30 min to ensure sufficient consumption. This time setting was enough for feed ingestion, and no adverse effect on the physiological state of the fish was observed due to oxygen consumption. After each feeding, the remaining feed pellets in each tank were counted to calculate the uneaten feed amount and siphoned out immediately. In the course of the feeding trial, tanks were checked daily for dead and moribund fish. The dead fish were removed from the tank immediately, while the moribund fish were dissected to rule out the presence of an obvious pathogenic organism after anesthetizing with an overdose of MS-222. Both the dead and moribund fish before dissection were weighed for calculation of the feed conversion ratio (FCR) at the final trial using the formula that FCR = Total feed consumption/[Final body weight + Weight of mortality (dead + moribund fish) − Initial body weight].

2.3. Water Collection and Metrics Determination

During the experiment, culture water was collected 2 h before the first feeding on days 14, 28, 42, and 70 of the feeding trial to measure the metrics, including ammonia nitrogen (NH₄⁺-N), nitrate nitrogen (NO₃⁻-N), nitrite (NO₂⁻-N), total phosphorus (TP), hardness, and pH. In detail, the values of water pH and dissolved oxygen (DO) were attained by using a portable water quality meter (WTW Multi 3630 IDS, Xylem Analytics Germany GmbH, Munich, Germany), and the determination of NH₄⁺-N, NO₃⁻-N, NO₂⁻-N, TP, and hardness in the aquaculture water was carried out according to the following Chinese standards: HJ535-2009, GB7493-87, HJ/T346-2007, HJ501-2009, GB11893-89, and GB7477-87 [30,31,32,33,34,35], accordingly.

2.4. Growth Performance

After the feeding trial, the fish were deprived of feed for 24 h. Then the fish in each tank were anesthetized by MS-222 with a dose of 90 mg/L (not lethal), followed by batch-weighing and counting the number to calculate the survival rate (SR) and final body weight (FBW) per fish. The feed intake (FI, g) of each fish in the feeding period was calculated according to the formula that FI = total feed consumption/[(initial number of fish + final number of fish)/2] [36]. The other metrics including SR (%), weight gain (WG, %), specific growth rate (SGR, %/d), and protein efficiency (PER, %) was calculated according to the formula as follows: SR = 100 × final number of fish/initial number of fish; WG = 100 × [FBW − IBW]/IBW; SGR =100 × [ln (FBW) − ln (IBW)]/d; PER = 100 × [(FBW)–(IBW)]/(feed protein intake).

2.5. Benefit-Cost Evaluation

This study employed the incremental feed cost for obtaining per kilogram of weight gain (IFC, USD/kg) to assess the economic viability of substituting SF with GKM in LMB aquaculture. The IFC was calculated as: IFC (USD/kg) = FCR × 1 kg × 0.15 × [(1 − A) × P_SF (USD/kg) + A × P_GKM (USD/kg) − P_SF (USD/kg)], where A is the ratio of SF replaced by GKM, P_SF was the price of SF, and P_GKM was the price of GKM. This economic model identified SF and GKM as primary cost drivers, while contributions from medical stone were considered negligible due to its low market price and little difference in usage amount among the groups. The prices of SF, GL, and KG used in this study were 439.04 USD/ton, 617.4 USD/ton, and 2469.6 USD/ton, respectively. As GKM is a mixture of GL and KG at a ratio of 2:1, the price of GKM per ton was 1234.8 USD/ton.

2.6. Proximate Composition, Amino Acids Profile, and Nutrient Retention Efficiency

At the onset of the feeding trial, three replications each containing 8 healthy fish (10.24 ± 0.193 g) that had fasted for 24 h were sampled from the acclimation tank and killed by an overdose of MS-222, and these were taken as the initial whole-body fish samples. Three fish per tank at the final feeding trial were collected and euthanized as the final whole-body fish samples. Whole-body fish samples were stored at −20 °C until the nutrient composition analysis. For analysis, sample processing was according to Cheng et al. [37]. The thawed fish samples in each group were cut and stirred quickly after being wiped dry. Part of them were immediately oven-dried at 105 °C to a constant weight for moisture determination, while the remaining samples were cooked at 120 °C for 30 min before oven-drying at 70 °C. The dried samples were finely ground into powder for crude protein (CP), lipid, and ash determination according to China National Standards (GB 5009.5-2016, GB 5009.6-2016, and GB 5009.4-2016) [38,39,40]. The phosphorus content was measured via inductively coupled plasma mass spectrometry (ICP-MS, ICAPTM RQ, Thermo Fisher Scientific, Waltham, Massachusetts, USA) according to the Chinese standard of GB 5009.87-2016 [41]. The energy was calculated according to the formula that gross energy (kJ/kg) = protein content (g/kg) × 23.9 kJ/g + lipid content (g/kg) × 37.8 kJ/g + starch content (g/kg) × 17.6 kJ/g [12]. To determine the amino acid profile, an approximately 300 mg sample was weighed into a hydrolysis tube and supplemented with 10 mL of HCl (6 mol/L). The tubes were then sealed with high-purity N_2, and the samples were digested at 110 °C for 22 h. After cooling the hydrolyzed product to room temperature (approximately 25 °C), it was diluted to a constant volume of 100 mL. Following, 1 mL diluted solution was pipetted and evaporated to dryness in a water bath at 40 °C. The dried samples were resuspended using 1 mL of HCl (0.02 mol/L) and filtered using a 0.22 μm syringe filter before analysis using a high-speed amino acid analyzer (L-8090, Hitachi High-Tech Science, Marunouchi, Tokyo, Japan). The nutrient retention efficiency (NRE) was calculated according to the equation as follows:

NRE_i (%) = 100 × (FBW × FBN_i − IBW × IBN_i)/(FI × DN_i)

where the FBW and IBW represent the final and initial fish body weight, respectively; and FBN_i, IBN_i, and DN_i represent the nutrients (such as crude protein, crude lipid, ash, energy, and amino acids) in the final fish body, initial fish body, and the feeding diets, respectively. The non-utilization amounts of nitrogen and phosphorus for obtaining per kilogram of weight gain (NUA, g/kg) were calculated according to the equations as follows:

NUA_i = 1000 × (FI × FC_i + IB × IBC_i − FB × FBC_i)/(FB − IB)

where i is the assessment indicator, referring to nitrogen and phosphorus; FI represents the feed intake (DM basis) during the experiment, and FC_i represents the content of nitrogen or phosphorus in the feed; IB and FB represent the initial and final biomass per tank, respectively; while IBC_i and FBC_i represent the content of nitrogen and phosphorus in the initial and final fish body, respectively.

2.7. Apparent Digestibility Coefficients

The apparent digestibility coefficients (ADC) were determined by using an exogenous standard method (Y₂O₃ as the indicator; Table 2) in this study. After the 10-week feeding trial, a continued 2-week feeding trial was conducted on 15 fish from the remaining fish in each tank to collect the digesta according to Zhang et al. [42]. Digesta collected per tank were pooled and stored at −20 °C before oven-drying to a constant weight at 70 °C [43]. The experimental feed and dried fecal samples (20–50 mg) were thoroughly ground, and the content of Yttrium was determined via inductively coupled plasma mass spectrometry (ICP-MS, ICAPTM RQ, Thermo Fisher Scientific, Waltham, Massachusetts, USA) according to the Chinese standard of HJ 766-2015 [44]. The ADC was calculated according to the equation as follows:

ADC_i = 100 × [1 − (FC_i × DC_Yttrium)/(DCi × FC_Yttrium)]

where FC_i and DC_i refer to the nutrient content in fish fecal and the providing diets, while DC_Yttrium and FC_Yttrium refer to the Yttrium content in feeding diets and fish fecal, respectively. All the contents mentioned above were on a DM basis.

2.8. Statistical Analysis

Data related to the SR, growth performance, nutrient composition, retention efficiency, ADCs, and water quality metrics were subjected to one-way ANOVA analysis using the software Origin 8.0 Pro SR4 (Origin Lab. Co., Ltd., Northampton, Massachusetts, USA) by using Tukey multiple range tests at the 5% level of significance. Results were expressed as means and standard deviation (SD). Regression models were used to identify the correlation between the dietary GKM inclusion levels and the determined indices when a significant difference was found by ANOVA. Linear or second-order polynomial regression models were applied when p < 0.05, and the second-order models were used if R² (Adj. R²) increased.

3. Results

3.1. Fish Survival Rate and Growth Performance

After the feeding trial, the results revealed that the GK05, GK10, and GK15 groups exhibited significantly higher SRs than the GK00 group, and no statistical differences were observed among GKM-substituted groups (GK05, GK10, and GK15). The growth performance metrics of LMB fed with different diets are summarized in

Table 3. Growth performance of largemouth bass Micropterus salmoides.

Items	GK00	GK05	GK10	GK15	p-Value
SR (%)	87.50 ± 1.966	98.15 ± 1.605	94.44 ± 2.775	94.44 ± 0.000	0.003
IBW (g)	10.34 ± 0.014	10.35 ± 0.030	10.47 ± 0.147	10.52 ± 0.181	0.328
FBW (g)	76.91 ± 0.658 b	86.61 ± 1.738 a	87.63 ± 3.334 a	78.88 ± 3.296 b	0.005
WG (%)	643.5 ± 5.35 b	736.8 ± 14.84 a	736.9 ± 25.96 a	649.7 ± 40.72 b	0.006
SGR (%/d)	2.87 ± 0.010 b	3.09 ± 0.105 a	3.03 ± 0.044 a	2.88 ± 0.077 b	0.022
FI (g)	87.8 ± 4.94 b	107.4 ± 3.94 a	108.7 ± 3.11 a	106.8 ± 4.23 a	0.003
FCR	1.32 ± 0.057	1.39 ± 0.103	1.43 ± 0.053	1.54 ± 0.053	0.053
PER (%)	1.53 ± 0.007	1.48 ± 0.091	1.39 ± 0.110	1.33 ± 0.087	0.146

Data are presented as mean ± standard deviation. n = 3. Different superscript letters in the same row indicated significant differences (p < 0.05, Tukey HSD test). IBW: initial body weight. FBW: final body weight. WG: weight gain. SGR: specific growth rate. FI: feed intake. FCR: feed conversion ratio. PER: protein efficiency ratio.

. The IBW had no significant differences across groups (p = 0.328). After the 10-week feeding trial, the GK05 and GK10 groups exhibited significant enhancements in FBW, WG, and SGR compared to the GK00 group, whereas the GK15 group displayed no growth-promoting effects relative to the GK00 group. FI in the GK05, GK10, and GK15 groups was significantly higher than that of the GK00 group (p = 0.003), though FCR and PER remained comparable across treatments (p = 0.053 and 0.146, respectively). Quadratic regression identified the optimal SF substitution levels with GKM based on WG (Adj. R² = 0.763) and FI (Adj. R² = 0.762) were 50.68% and 68.01%, respectively (Figure 1).

3.2. Economic Benefit Evaluation

The economic analysis revealed feed requirements of diets GK00, GK05, GK10, and GK15 to obtain one kilogram of WG were 1.32 kg, 1.39 kg, 1.43 kg, and 1.54 kg, respectively (

Table 4. Economic benefit evaluation based on feed cost of Micropterus salmoides.

Items	GK00	GK05	GK10	GK15
Usage amount of diets (kg/kg)	1.32 ± 0.057	1.39 ± 0.102	1.43 ± 0.053	1.54 ± 0.053
Usage amount of SF (g/kg)	198 ± 8.5	139 ± 10.3	72 ± 2.6	0
Usage amount of GKM (g/kg)	0	70 ± 5.1	143 ± 5.3	231 ± 7.9
IFC (USD/kg)	-	0.06 ± 0.011 c	0.12 ± 0.008 b	0.20 ± 0.010 a

Data are presented as mean ± standard deviation. n = 3. Values in the same row with different superscript letters indicated significant differences (p < 0.05, Tukey HSD test). IFC: the incremental feed cost for obtaining per kilogram of weight gain.

). The IFC of the GK05, GK10, and GK15 was 0.06 (USD/kg), 0.12 (USD/kg), and 0.20 (USD/kg), respectively. Although a significantly increased IFC was observed with the elevated dietary GKM inclusion level (p < 0.001), the differences seemed to be within an acceptable range.

3.3. Proximate Composition of Fish Body and Nutrition Retention Efficiency

Whole-body proximate composition analysis indicated no significant differences in moisture (p = 0.375), protein (p = 0.405), and energy content (p = 0.194) across the groups (

Table 5. Effect of feeding diets on fish body proximate composition and nutrition retention efficiency of Micropterus salmoides.

Items	GK000	GK05	GK10	GK15	p-Value	Regression Model	Adj. R2
Moisture (%)	70.25 ± 0.353	69.96 ± 0.109	70.82 ± 0.143	70.03 ± 0.352	0.375	-	-
Crude protein (%)	18.75 ± 0.531	18.55 ± 0.443	18.05 ± 0.576	18.80 ± 0.409	0.405	-	-
Crude lipid (%)	6.08 ± 0.024 a	5.91 ± 0.070 a	5.62 ± 0.001 b	5.51 ± 0.114 b	<0.001	6.093 − 0.007 x + 1.106 e−5 x2	0.885
Ash (%)	4.49 ± 0.004 b	4.87 ± 0.241 ab	5.03 ± 0.209 ab	5.23 ± 0.245 a	0.047	4.566 + 0.007 x	0.637
Phosphorus (%)	0.32 ± 0.001 b	0.35 ± 0.027 ab	0.38 ± 0.020 a	0.38 ± 0.023 a	0.019	0.316 + 0.008 x − 2.638 e−4 x2	0.620
Gross energy (MJ/kg)	6.78 ± 0.118	6.58 ± 0.109	6.44 ± 0.137	6.63 ± 0.164	0.194	-	-
PRE (%)	31.36 ± 0.458 a	28.48 ± 1.488 ab	28.44 ± 1.420 ab	27.34 ± 0.416 b	0.034	31.125 − 0.080 x + 4.367 e−4 x2	0.579
LRE (%)	46.03 ± 1.013 a	45.37 ±1.492 a	42.74 ± 0.518 ab	38.06 ± 1.030 b	0.002	46.029 + 0.011 x − 9.023 e−4 x2	0.867
ARE (%)	21.27 ± 1.013	20.99 ± 1.839	21.98 ± 1.322	19.86 ± 1.095	0.461	-	-
TPRE (%)	42.06 ± 1.655 b	44.41 ± 1.246 ab	46.30 ± 2.380 ab	48.26 ± 1.242 a	0.017	42.225 + 0.061 x	0.548
EER (%)	28.72 ± 0.788 a	26.49 ± 0.462 ab	26.35 ± 0.979 ab	24.79 ± 0.660 b	0.005	28.255 − 0.035 x	0.751

Data are presented as mean ± standard deviation. n = 3. Different superscript letters in the same row indicated significant differences among the groups (p < 0.05, Tukey HSD test). The moisture, protein, lipid, ash, and energy content in the initial fish body were 74.24 ± 0.252%, 16.70 ± 0.246%, 3.73 ± 0.0435%, 4.44 ± 0.044%, and 5.40 ± 0.075%, accordingly. PRE: protein retention efficiency. LPE: lipid retention efficiency. ARE: Ash retention efficiency. TPRE: total phosphorus retention efficiency. ERE: energy retention efficiency.

). Crude lipid content was significantly reduced in the GK10 and GK15 groups compared to the GK00 group, whereas the GK05 group showed no statistically significant difference from that of the GK00 group. Ash content remained comparable between the GK05, GK10, and GK00 groups, but demonstrated a notable elevation in the GK15 group relative to the GK00 group. Phosphorus levels displayed no variation between the GK05 and GK00 groups, yet increased significantly in both the GK10 and GK15 groups compared to the GK00 group. Furthermore, a quadratic increase of phosphorus content (Adj. R² = 0.620), a linear increase of ash content (Adj. R² = 0.637), and a quadratic decline of lipid content (Adj. R² = 0.885) in the final fish body as the increase in substitution ratio of SF by GKM was found. Regarding nutrient retention efficiency, the GK15 group displayed significant reductions in protein, lipid, and energy retention efficiency relative to the GK00 group, while those of the GK05 and GK10 groups remained comparable to the GK00 group. In contrast, phosphorus retention efficiency demonstrated an inverse trend, with GK15 showing a significant elevation compared to the GK00 group, whereas no significant differences were observed among GK05, GK10, and GK00 groups. Fitting models further revealed quadratic decreases in protein (Adj. R² = 0.579) and lipid retention efficiency (Adj. R² = 0.867), accompanied by a linear reduction in energy retention efficiency (Adj. R² = 0.751) as GKM substitution ratios increased. The phosphorus retention efficiency presented a linear increase as the dietary GKM inclusion levels increased.

3.4. Hydrolyzed Amino Acids Profile of the Fish Body and Nutrient Retention Efficiency

Table 6. Hydrolyzed amino acids profile in the whole body of Micropterus salmoides (%, DM basis).

Items	GK00	GK05	GK10	GK15	p-Value
Val	2.97 ± 0.147	2.85 ± 0.086	2.84 ± 0.049	2.90 ± 0.148	0.532
Lys	5.16 ± 0.438	4.81 ± 0.096	4.79 ± 0.131	4.91 ± 0.222	0.334
Ile	2.71 ± 0.210	2.51 ± 0.080	2.51 ± 0.057	2.56 ± 0.112	0.281
Leu	4.56 ± 0.373	4.22 ± 0.094	4.26 ± 0.104	4.33 ± 0.203	0.327
Phe	2.54 ± 0.091	2.40 ± 0.066	2.43 ± 0.050	2.46 ± 0.095	0.207
Met	1.74 ± 0.103	1.66 ± 0.023	1.64 ± 0.037	1.68 ± 0.056	0.309
Thr	2.73 ± 0.085	2.60 ± 0.058	2.63 ± 0.049	2.67 ± 0.113	0.322
Arg	3.92 ± 0.248	4.02 ± 0.081	3.96 ± 0.075	4.06 ± 0.154	0.682
Cys	0.60 ± 0.088	0.52 ± 0.005	0.53 ± 0.090	0.54 ± 0.04	0.481
His	1.39 ± 0.086	1.30 ± 0.055	1.28 ± 0.043	1.33 ± 0.087	0.350
Asp	5.93 ± 0.274	5.62 ± 0.130	5.70 ± 0.094	5.75 ± 0.244	0.342
Ser	2.66 ± 0.103	2.63 ± 0.046	2.65 ± 0.053	2.68 ± 0.110	0.890
Glu	9.17 ± 0.311	8.84 ± 0.188	8.86 ± 0.181	9.04 ± 0.331	0.419
Gly	4.73 ± 1.595	5.76 ± 0.162	5.46 ± 0.111	5.72 ± 0.100	0.420
Ala	4.18 ± 0.422	4.45 ± 0.108	4.32 ± 0.067	4.49 ± 0.112	0.391
Tyr	1.91 ± 0.181	1.68 ± 0.027	1.68 ± 0.035	1.72 ± 0.127	0.102
Pro	2.21 ± 0.715	2.63 ± 0.173	2.60 ± 0.052	2.61 ± 0.058	0.491
EAA	22.41 ± 1.433	21.05 ± 0.491	21.11 ± 0.463	21.51 ± 0.942	0.317
NEAA	36.70 ± 2.472	37.46 ± 0.877	37.04 ± 0.533	37.95 ± 1.198	0.752
TAA	59.10 ± 1.631	58.51 ± 1.356	58.14 ± 0.993	59.46 ± 2.139	0.750

Data are presented as mean ± standard deviation. n = 3. EAA: essential amino acid. NEAA: non-essential amino acid. TAA: total amino acid.

shows no significant effect of feeding diets on the fish body hydrolyzed amino acid profile. The contents of total amino acid (TAA), essential amino acid (EAA), and non-essential amino acid (NEAA) of the fish body were about 50%, 22%, and 37%, respectively. The highest and lowest content of essential amino acids were Lys (4.79–5.16%) and Met (1.64–1.74%), respectively.

The dietary GKM inclusion significantly decreased the retention efficiency of TAA (p = 0.005), EAA (p = 0.007), NEAA (p = 0.005), Ile (p = 0.004), Leu (p = 0.015), Phe (p = 0.001), Met (p = 0.002), Thr (p = 0.002), Arg (p = 0.007) and Asp (p = 0.001), while no impaired on retention efficiency of His (p = 0.088), Cys (p = 0.297), Glu (p = 0.156), and Pro (p = 0.190). Retention efficiency of Val, Ser, Gly, and Ala in the GK10 and GK15 groups was significantly lower than that of the GK00 group, while those in the GK05 group were no different from the GK05 group. The Lys retention efficiency in the GK10 group was significantly lower than that of the GK00 group, and there was no difference between the GK00, GK05, and GK15 groups. The Tyr retention efficiency in the GK05 and GK10 groups was significantly lower than that of the GK00 group, and there was no difference between the GK00 and GK15 groups (

Table 7. Effect of feeding diets on the amino acid retention efficiency of Micropterus salmoides.

Items	GK00	GK05	GK10	GK15	p-Value	Regression Model	Adj. R2
Val (%)	28.96 ± 0.414 a	27.61 ± 0.228 ab	26.40 ± 0.933 b	26.00 ± 1.252 b	0.008	28.994 − 0.3433 x + 0.00942 x2	0.699
Lys (%)	36.30 ± 0.566 a	33.38 ± 0.287 ab	32.95 ± 1.658 b	33.43 ± 1.411 ab	0.024	36.219 − 0.6914 x + 0.03405 x2	0.758
Ile (%)	28.10 ± 0.638 a	26.02 ± 0.182 b	25.36 ± 0.870 b	25.21 ± 1.008 b	0.004	28.049 − 0.4753 x + 0.01930 x2	0.745
Leu (%)	29.21 ± 0.975 a	26.54 ± 0.172 b	26.43 ± 1.271 b	26.42 ± 1.162 b	0.015	29.085 − 0.5687 x + 0.02661 x2	0.600
Phe (%)	27.38 ± 0.056 a	24.38 ± 0.114 b	24.44 ± 1.009 b	24.47 ± 0.837 b	0.001	27.227 − 0.6282 x + 0.03032 x2	0.740
Met (%)	36.57 ± 0.154 a	32.89 ± 0.940 b	32.20 ± 1.336 b	33.74 ± 0.956 b	0.002	36.528 − 0.9650 x + 0.05212 x2	0.779
Thr (%)	31.93 ± 0.202 a	29.12 ± 0.272 b	28.68 ± 1.204 b	28.14 ± 1.098 b	0.002	31.811 − 0.5789 x + 0.02284 x2	0.749
His (%)	24.40 ± 0.051	22.49 ± 0.628	22.12 ± 1.384	23.92 ± 1.531	0.088	-	-
Arg (%)	32.80 ± 0.117 a	30.48 ± 0.363 b	29.43 ± 1.320 b	30.28 ± 1.001 b	0.007	32.829 − 0.6467 x + 0.03164 x2	0.709
Cys (%)	28.01 ± 7.068	22.82 ± 0.918	23.39 ± 3.919	25.84 ± 2.394	0.297	-	-
ASP (%)	30.71 ± 0.631 a	27.99 ± 0.191 b	27.47 ± 1.081 b	26.66 ± 1.038 b	0.001	30.590 − 0.5395 x + 0.01904 x2	0.780
Ser (%)	29.99 ± 0.302 a	27.88 ± 0.322 ab	27.28 ± 1.241 b	27.09 ± 1.034 b	0.010	29.934 − 0.4736 x + 0.01919 x2	0.674
Glu (%)	28.13 ± 0.238	26.87 ± 0.195	26.73 ± 1.120	27.39 ± 0.888	0.156	-	-
Gly (%)	50.27 ± 2.707 a	48.83 ± 0.907 ab	44.13 ± 2.026 c	45.78 ± 0.799 bc	0.004	50.746 − 0.8252 x + 0.03082 x2	0.555
Ala (%)	38.29 ± 0.914 a	36.65 ± 0.394 a	33.59 ± 1.291 b	34.19 ± 0.615 b	0.000	38.546 − 0.6438 x + 0.02245 x2	0.770
Tyr (%)	29.42 ± 0.519 a	24.73 ± 0.343 b	24.67 ± 0.878 b	27.26 ± 2.051 ab	0.003	29.319 − 1.2235 x + 0.07287 x2	0.764
Pro (%)	31.32 ± 0.553	29.50 ± 2.093	29.24 ± 0.743	30.35 ± 0.438	0.190	-	-
EAA (%)	30.97 ± 0.525 a	28.41 ± 0.199 b	27.94 ± 1.199 b	27.97 ± 1.125 b	0.007	30.894 − 0.5793 x + 0.02597 x2	0.696
NEAA (%)	32.44 ± 0.419 a	30.41 ± 0.424 b	29.41 ± 1.098 b	30.13 ± 0.777 b	0.005	32.478 − 0.5720 x + 0.02755 x2	0.734
TAA (%)	31.90 ± 0.068 a	29.66 ± 0.338 b	28.86 ± 1.131 b	29.31 ± 0.908 b	0.005	31.888 − 0.5747 x + 0.02691 x2	0.739

Data are presented as mean ± standard deviation. n = 3. Different superscript letters in the same row indicated significant differences among the GK00, GK05, GK10, and GK15 groups (p < 0.05, Tukey HSD test). EAA: essential amino acid. NEAA: non-essential amino acid. TAA: total amino acid.

).

3.5. Apparent Digestibility Coefficient

As detailed in

Table 8. Effect of feeding diets on the apparent digestibility coefficient of Micropterus salmoides.

Items	GK00	GK05	GK10	GK15	p-Value	Regression Model	Adj. R2
Dry matter (%)	54.64 ± 0.383 a	65.64 ± 4.102 a	58.82 ± 4.704 a	37.63 ± 6.790 b	<0.001	54.810 + 0.551 x − 0.007 x2	0.855
Crude protein (%)	87.08 ± 0.690 a	89.99 ± 1.449 a	83.81 ± 5.098 a	72.47 ± 4.756 b	0.004	87.362 + 0.168 x − 0.003 x2	0.789
Crude lipid (%)	89.78 ± 1.534 ab	92.21 ± 1.904 a	92.38 ± 0.858 a	81.04 ± 7.554 b	0.026	89.322 + 1.543 x − 0.138 x2	0.635
TP (%)	39.57 ± 3.999 b	40.99 ± 0.504 b	53.04 ± 4.127 ab	59.70 ± 6.199 a	0.010	36.816 + 0.226 x	0.700
Val (%)	91.13 ± 1.697 a	92.98 ± 0.492 a	91.47 ± 1.483 a	83.17 ± 1.637 b	0.000	90.955 + 1.015 x − 0.102 x2	0.888
Lys (%)	95.62 ± 1.294 a	95.66 ± 0.284 a	92.11 ± 2.427 a	85.41 ± 4.214 b	0.003	95.646 + 0.327 x − 0.067 x2	0.761
Ile (%)	93.45 ± 0.404 a	95.06 ± 1.372 a	91.90 ± 1.483 a	84.70 ± 1.436 b	0.000	93.483 + 0.734 x − 0.088 x2	0.923
Leu (%)	95.16 ± 0.319 a	95.75 ± 0.720 a	93.17 ± 1.375 a	87.14 ± 1.509 b	0.000	95.145 + 0.461 x − 0.066 x2	0.921
Phe (%)	94.22 ± 0.426 a	95.29 ± 0.467 a	92.08 ± 2.390 a	86.04 ± 3.249 b	0.002	94.293 + 0.511 x − 0.071 x2	0.780
Met (%)	80.06 ± 5.047 a	81.82 ± 8.802 a	67.61 ± 4.467 ab	55.16 ± 4.976 b	0.002	80.948 + 0.353 x − 0.142 x2	0.751
Thr (%)	90.98 ± 1.462 a	93.55 ± 0.301 a	91.14 ± 1.792 a	83.26 ± 2.455 b	0.000	90.953 + 1.057 x − 0.105 x2	0.862
His (%)	96.27 ± 0.754 a	96.88 ± 0.412 a	95.37 ± 1.470 a	91.31 ± 2.058 b	0.004	96.252 + 0.371 x − 0.047 x2	0.753
Arg (%)	96.54 ± 0.434 ab	97.34 ± 0.159 a	96.79 ± 0.905 ab	94.00 ± 2.102 b	0.033	96.494 + 0.376 x − 0.036 x2	0.562
Cys (%)	87.40 ± 0.945 ab	89.90 ± 0.944 a	84.16 ± 2.218 b	68.34 ± 1.991 c	0.000	87.817 + 1.927 x − 0.217 x2	0.891
Asp (%)	90.04 ± 2.385 a	92.77 ± 0.121 a	89.35 ± 3.004 a	80.66 ± 3.279 b	0.002	90.080 + 1.083 x − 0.114 x2	0.791
Ser (%)	93.04 ± 0.738 a	94.95 ± 0.210 a	93.19 ± 1.528 a	87.34 ± 2.520 b	0.001	93.022 + 0.785 x − 0.078 x2	0.806
Glu (%)	92.74 ± 1.128 a	95.05 ± 0.175 a	92.22 ± 2.719 a	85.14 ± 2.743 b	0.002	92.784 + 0.896 x − 0.094 x2	0.798
Gly (%)	87.35 ± 1.479 a	91.58 ± 0.399 a	89.96 ± 2.685 a	80.42 ± 1.021 b	0.000	87.249 + 1.618 x − 0.138 x2	0.889
Ala (%)	90.62 ± 2.157 a	93.30± 0.295 a	91.78 ± 2.457 a	83.59 ± 3.416 b	0.005	90.492 + 1.178 x − 0.109 x2	0.735
Cyr (%)	94.17 ± 0.891 a	96.79 ± 1.275 a	93.80 ± 3.270 a	87.17 ± 1.985 b	0.003	94.272 + 0.908 x − 0.093 x2	0.771
Pro (%)	91.06 ± 1.133 b	93.85 ± 0.058 a	92.77 ± 0.996 ab	84.54 ± 0.287 c	0.000	90.897 + 1.239 x − 0.110 x2	0.952
EAA (%)	92.59 ± 0.494 a	93.96 ± 0.703 a	90.42 ± 1.825 a	82.33 ± 3.191 b	0.000	92.611 + 0.731 x − 0.094 x2	0.872
NEAA (%)	91.31 ± 1.380 a	93.98 ± 0.131 a	91.57 ± 2.477 a	83.74 ± 2.362 b	0.000	91.290 + 1.074 x − 0.105 x2	0.837
TAA (%)	92.34 ± 0.908 a	94.30± 0.240 a	91.55 ± 2.054 a	84.07 ± 2.545 b	0.001	92.336 + 0.865 x − 0.094 x2	0.861

Data are presented as mean ± standard deviation. n = 3. Different superscript letters in the same column indicated significant differences among the GK00, GK05, GK10, and GK15 groups (p < 0.05, Tukey HSD test). TP: total phosphorus. EAA: essential amino acid. NEAA: non-essential amino acid. TAA: total amino acid.

, dietary GKM inclusion levels demonstrated significant dose-dependent effects on nutrient ADCs. The results showed that there was no adverse effect on the digestibility of DM, CP, CF, TAA, EAA, NEAA, and the specific amino acids when one-third or two-thirds SF was replaced by the GKM. However, the digestibility of those nutrients was significantly decreased when 100% SF was replaced by GKM. The digestibility of phosphorus in the GK05 and GK10 groups remained comparable to that of the GK00 group, whereas the GK15 group exhibited a significant elevation versus that of the GK00 group. Furthermore, regression modeling revealed quadratic declines in all aforementioned ADC parameters except phosphorus with increasing GKM inclusion levels. In contrast, the ADC of phosphorus demonstrated a strong linear increase corresponding to the incremental GKM substitution rates.

3.6. Water Quality Metrics

During the feeding trial, the substitution of SF with GKM did not significantly alter the pH and hardness of culture water (p > 0.05,

Table 9. Effect of feeding diets on the culture water quality of Micropterus salmoides.

Metrics	Days	GK00	GK05	GK10	GK15	p-Value	Regression Model	Adj. R2
pH	14	7.38 ± 0.028	7.42 ± 0.056	7.45 ± 0.110	7.46 ± 0.066	0.682	-	-
	28	7.60 ± 0.120	7.59 ± 0.021	7.60 ± 0.040	7.62 ± 0.058	0.894	-	-
	42	7.29 ± 0.078	7.26 ± 0.058	7.23 ± 0.036	7.31 ± 0.076	0.525	-	-
	70	6.97 ± 0.127	6.76 ± 0.068	6.83 ± 0.171	6.74 ± 0.137	0.299	-	-
NH4+-N (mg/L)	14	7.61 ± 0.594 a	0.33 ± 0.146 b	0.65 ± 0.339 b	0.32 ± 0.09 b	<0.001	7.1125 − 0.2285 x + 0.0016 x2	0.866
	28	0.31 ± 0.008 b	1.12 ± 0.04 a	0.70 ± 0.25 ab	0.61 ± 0.090 ab	0.023	0.3971 + 0.0196 x − 1.8378 x2	0.608
	42	2.33 ± 0.283 a	2.01 ± 0.355 a	0.98 ± 0.109 b	1.86 ± 0.040 a	0.001	2.5229 − 0.0365 x + 2.8611 e−4 x2	0.440
	70	0.68 ± 0.099	0.74 ± 0.133	0.74 ± 0.198	0.59 ± 0.189	0.690	-	-
NO2−-N (mg/L)	14	0.95 ± 0.057 b	1.70 ± 0.212 a	2.06 ± 0.007 a	1.07 ± 0.087 b	<0.001	0.9020 + 0.0414 x − 3.9257 e−4 x2	0.888
	28	0.12 ± 0.014	0.28 ± 0.130	0.23 ± 0.101	0.31 ± 0.142	0.392	-	-
	42	0.96 ± 0.311	1.11 ± 0.098	0.77 ± 0.098	1.16 ± 0.231	0.131	-	-
	70	0.32 ± 0.092	0.34 ± 0.127	0.29 ± 0.035	0.46 ± 0.150	0.441	-	-
NO3−-N (mg/L)	14	1.23 ± 0.106	1.63 ± 0.735	1.47 ± 0.255	1.39 ± 0.346	0.818	-	-
	28	9.86 ± 0.502 c	14.85 ± 0.670 a	14.56 ± 0.437 ab	13.29 ± 0.307 b	<0.001	10.1698 + 0.1677 x − 0.0014 x2	0.852
	42	9.56 ± 0.269 b	13.38 ± 2.337 ab	14.33 ± 0.766 a	12.11 ± 0.946 ab	0.036	9.53389 + 0.1621 x − 0.0014 x2	0.603
	70	11.09 ± 2.022	13.60 ± 2.088	14.39 ± 2.640	14.28 ± 1.616	0.392	-	-
Hardness (mg/L)	14	98.8 ± 2.13	102.3 ± 7.23	103.6 ± 6.13	105.7 ± 6.13	0.675	-	-
	28	126.9 ± 4.96	131.4 ± 2.65	132.1 ± 7.04	135.7 ± 3.23	0.320	-	-
	42	121.4 ± 1.42	126.0 ± 1.53	125.7 ± 4.18	129.4 ± 1.74	0.066	-	-
	70	94.3 ± 0.04	97.0 ± 1.15	91.6 ± 12.3	97.6 ± 13.4	0.676	-	-
TP (mg/L)	14	1.06 ± 0.233 ab	1.25 ± 0.046 ab	1.40 ± 0.064 a	0.88 ± 0.210 b	0.017	1.0175 + 0.0153 x − 1.6377 e−4 x2	0.560
	28	0.57 ± 0.113 c	1.00 ± 0.102 b	1.62 ± 0.086 a	1.64 ± 0.068 a	<0.001	0.5140 + 0.0212 x − 9.5750 e−5 x2	0.912
	42	0.72 ± 0.007 d	1.12 ± 0.085 c	1.84 ± 0.072 b	2.11 ± 0.093 a	<0.001	0.6589 + 0.0184 x − 3.5350 e−5 x2	0.949
	70	0.66 ± 0.050 c	1.27 ± 0.156 b	1.87 ± 0.106 a	1.88 ± 0.280 a	<0.001	0.6228 + 0.0251 x − 1.1009 e−4 x2	0.952

Data presented as mean ± standard error (mean ± SD). Different superscript symbols in the same row indicate significant differences among treatments by one-way ANOVA analysis using the Tukey test at the level of p < 0.05. TP, total phosphorus; n = 3.

). In week 2 of the experiment, the substitution of SF with GKM significantly reduced the NH₄⁺-N content in the water (p < 0.001). As the proportion of GKM increased, the NH₄⁺-N content in the water showed a quadratic decreasing trend (Adj. R² = 0.866). However, this pattern was reversed in the fourth week of the culture experiment, exhibiting a quadratic elevation of the NH₄⁺-N content in the water (Adj. R² = 0.608), where substitution of one-third SF significantly increased the NH₄⁺-N content compared to the control group. Measurements in week 6 of the experiment revealed no significant difference in the NH₄⁺-N content among the GK05, GK15, and GK00 groups, whereas a significant decrease in the NH₄⁺-N content was observed in the GK10 group. In week 10 of the experiment, there was no significant difference in the NH₄⁺-N content across the groups (p = 0.690). In week 2 of the experiment, the substitution of one-third or two-thirds SF with GKM significantly increased the NO₂⁻-N content in the water, while the substitution of 100% SF had no significant effect on the NO₂⁻-N content. In the fourth, sixth, and tenth weeks, there was no significant difference in the NO₂⁻-N content between the groups with or without GKM inclusion. In the second week, the substitution of SF with GKM had no significant effect on the NO₃⁻-N content, but in the fourth week, the substitution of SF with GKM significantly increased the NO₃⁻-N content. In the sixth week, the substitution of one-third or 100% SF with GKM had no significant effect on the NO₃⁻-N content, while the substitution of two-thirds SF significantly increased its content in water. In the tenth week, there was no significant difference in the NO₃⁻-N content between the GKM substitution groups and the control group. However, TP concentrations showed persistent increases across experimental phases. Whole substitutions of SF by GKM elevated the TP content in week 2, with all GKM inclusion groups maintaining significantly higher TP than those of the control group from week 4 through trial termination. Quadratic regression confirmed a dose-dependent upward trend in water TP content with increasing dietary GKM inclusion.

3.7. The Non-Utilization Amounts of Nitrogen and Phosphorus

Figure 2 illustrates the impacts of graded SF replacement with GKM on the NUA_N and NUA_P. The results showed partial substitution (one-third or two-thirds) of SF by GKM demonstrated comparable NUA_N values to those of the GK00 group, whereas complete (100%) SF replacement significantly enhanced the NUAN. Conversely, both one-thirds and two-thirds GKM substitutions markedly reduced NUAP relative to the control group, though no significant difference emerged between the GK15 and GK00 groups.

4. Discussion

4.1. Effect on SR, Growth Performance, and Feed Utilization

Macroalgae-derived ingredients such as GL sharing comparable protein content (3–30% in DM) with the SF, exhibited fundamentally distinct carbohydrate compositions [45,46]. The main carbohydrate type in SF was starch, while those of the GKM were mainly NSPs, such as carrageenan and agar in Gracilaria spp. and konjac Glucomannan [17,18]. Previous studies showed carnivorous fish such as LMB were limited in metabolic capacity for starch [47,48,49], and significant variability in utilization efficiency across carbohydrate types [50,51,52]. In this study, iso-nitrogenous and iso-lipidic diets with close content and different compositions of carbohydrates were formulated by replacing SF with graded levels (0–15%) of GKM. The results revealed that GKM inclusion significantly improved the SR of LMB, aligning with findings in Oncorhynchus mykiss [53,54] and Pelteobagrus fulvidraco [55]. A meta-analysis by Thépot et al. further demonstrated a 33% average SR improvement in pathogen-challenge models through macroalgae-mediated immunomodulation, involving enhanced immune defense and homeostasis regulation [56].

The results of this study also showed partial SF replacement with GKM (5–10% inclusion) could significantly enhance fish growth (FBW, WG, and SGR), consistent with the findings of Kong et al., who observed an improved WG in LMB by feeding the diets with kelp meal (5–9%) inclusion [16]. However, complete SF replacement (15% GKM) yielded no additional benefits on growth-prompting, suggesting an optimal substitution threshold of SF by GKM. This dose-dependent growth response to GKM in the current study was paralleled by observations of algal effects on growth performance in Mugil liza [57], Carassius auratus [58], Oreochromis niloticus [59], Labeo rohita [60], and hybrid snakehead [15]. The absence of growth enhancement in fish that fed high GKM-supplemented diets could be attributable to the progressive accumulation of NSPs in formulated feeds, which exhibit dose-dependent anti-nutritional effects [61]. The study conducted by Fernandes et al. may exemplify this viewpoint to some extent, in which the content of macroalgae supplementation with no impairment on fish growth was significantly evaluated by disrupting the structural polysaccharides through the physical, chemical, and biological pretreatment technologies [62]. In this study, the substitution of SF with GKM significantly increased the FI of LMB. This may be due to the lower or slower energy supply of the diets with GKM inclusion compared to those of the GK00 group, necessitating a higher FI to compensate for the energy shortfall resulting from the feed consumption [63]. However, the FCR and PER were not impaired by the dietary overdose of GKM, suggesting a synergistically enhanced growth while maintaining the feed efficiency of dietary GKM.

4.2. Effect on ADC

Previous studies have shown that supplementation of an appropriate amount of macroalgae to fish feed has no negative effect on (and even improved) the ADC of feed nutrients, while an overdose significantly reduced the digestibility of feed nutrients, further influencing the feed utilization and fish growth [64,65,66,67,68]. In this study, partial substitution of SF with GKM (33.3% or 66.7% substitution) did not affect the ADC of DM, protein, lipid, and amino acids in the feed of LMB, while a higher substitution ratio (15%) significantly reduced the ADC of these nutrients. This may be related to the corresponding increase in dietary NSP content as the substitution ratio increases. Excessive NSP in feed accelerated the passage of ingested nutrients through the digestive tract, resulting in a decreased absorption rate of nutrients [69,70]. The reduction in ADC_CP caused by high NSP feed may relate to the stimulation of NSP on intestinal mucus secretion in fish, exacerbating the nitrogen loss [71]. Furthermore, it has been suggested that poor ADC_CP in high macroalgae supplementation diets for fish was due to poor digestion of the algal-derived proteins and limited capacity to hydrolyze complex polysaccharides [72]. Furthermore, Leenhouwers et al. proposed that NSP negatively impacted the ADC_CP due to the impairment of mineral or water absorption and the elevated viscosity of intestinal digesta in fish [73]. The process of protein digestion, absorption, and utilization in fish is essentially an amino acid turnover process. Therefore, the mechanism by which high levels of NSP impair protein digestibility may analogously apply to amino acid digestibility, though this proposition requires further verification.

Notably, this study revealed a dose-dependent enhancement in ADC_TP with the incremental dietary incorporation of GKM. The finding corresponded with the study by Wang et al., which documented an improved ADC_TP in LMB that fed diets supplemented with 6–9% enzymatic-hydrolyzed kelp meal [74]. Similarly, Zhang et al. observed an elevation in ADC_TP with graded levels (0–10%) of NSP supplementation in the diet for yellowtail kingfish (Seriola lalandi), though with maintained protein and ash digestibility [75]. The improvement in ADC_TP from macroalgae supplementation is likely attributable to the superior bioavailability of macroalgal phosphorus compared to that in the SF, in which phytate-phosphorus was predominant. However, phytate-phosphorus exhibits limited digestibility in fish due to their inherently low endogenous phytase activity. The distinctive polysaccharide matrix in macroalgae may promote phosphorus release through ion exchange mechanisms during gastrointestinal digestion.

4.3. Effect on Body Composition and Nutrient Retention Efficiency

In this study, the moisture content of LMB was 70–72%, protein content was 18–19%, lipid content was 4.7–6.1%, ash content was 4.4–5.7%, and gross energy was 6–7 MJ/kg. These values were consistent with previous research on LMB nutritional profiles [16,76,77,78]. However, the protein and lipid contents of LMB at the end of the feeding trial in this study were higher than those reported by Wu et al. [79], and this discrepancy may be potentially attributable to the difference in fish size. Generally, juvenile fish typically demonstrated higher moisture content and lower protein and lipid content compared to adult fish. The fish size in this study was close to 70 g at the final feeding trial, obviously larger than that (0.7 g) used in the study conducted by Wu et al. [79].

This study found that the dietary inclusion of 0–15% macroalgae had no significant effect on the protein content but significantly reduced the lipid content of LMB. However, Yang et al. discovered that adding 5% GL to the LMB diet did not significantly affect either protein or lipid content [80]. Kong et al. observed that 0–9% kelp powder supplementation significantly increased the protein content of LMB while not significantly affecting the lipid content [16]. Comparative analysis of feed formulations across the three studies revealed that despite all employing the iso-nitrogenous and iso-lipidic principles, significant variations existed in dietary protein and lipid levels, macronutrient ratios (such as protein to lipid and carbohydrate to lipid ratios), lipid ingredients combinations (fish oil, soybean oil, or their combination), and macroalgal nutritional profiles, all potentially affecting fish proximate composition [81,82,83,84,85,86]. Furthermore, previous studies have indicated that high-starch feed induced lipid accumulation in farmed fish [87,88], while NSP supplementation significantly reduced the lipid deposition in LMB caused by ingestion of high-starch feed [61,89], which partly corroborated the reliability of the findings in this study.

Investigating NRE in farmed fish could provide critical insights for optimizing feed formulations, minimizing nutritional waste, reducing production costs, and mitigating aquatic pollution in aquaculture systems. The discrepancy in NRE_i among the groups in this study suggested the different influences on feed nutrient efficiency of dietary varying levels of GKM. The decreased trend of protein, lipid, and energy retention efficiency may be related to the significant elevation in FI with the increased dietary GKM levels, as inferred from the calculation equation for NRE. Furthermore, the lower retention efficiency of amino acids except His, Cys, Glu, and Pro was found when feeding the diet with higher GKM inclusion levels, suggesting a relative lack of amino acids in the corresponding diets [90]. This phenomenon may be attributed to the significantly reduced amino acid digestibility observed in diets with high GKM inclusion levels, despite comparable amino acid composition profiles among experimental feed formulations. The compromised amino acid utilization efficiency under high GKM supplementation likely reflects structural interference of algal polysaccharides with proteolytic enzyme activity, as evidenced by Sotoudeh and Mardani [91]. Crucially, GKM substitutions enhanced the NRE_TP, consistent with the results that improved phosphorus digestibility and elevated body phosphorus content in LMB of this study. These findings suggested that exogenous inorganic phosphorus supplementation could be strategically reduced in macroalgae-enriched feeds to advance environmentally sustainable feed development.

4.4. Effect on Culture Water Quality

In aquaculture systems, feed composition and feed utilization efficiency exert a pivotal role in water quality dynamics via the regulation of metabolic byproduct flux [92,93,94]. Recent years have witnessed a growing interest in the application of macroalgae in aquafeeds, with particular attention to their impacts on water quality. Studies indicated that nitrogenous compounds, phosphorus, and organic carbon from uneaten feed constitute major pollutants in aquaculture effluents [95], and their emission levels are directly linked to dietary nutrient utilization efficiency [96]. This study revealed that partial substitutions of SF with GKM increased the nitrogen and phosphorus ingestion while simultaneously improving fish growth, leading to elevated retention of nitrogen and phosphorus in fish bodies, all of these resulting in a comparative NUA_N to the control group. Conversely, 100% SF substitution significantly increased the nitrogen and phosphorus ingestion while not increasing the nitrogen/phosphorus accumulation in the fish body, thereby elevating nutrient emissions. The finding in this study was consistent with the research conducted by Jørgensen et al., wherein high dietary NSP compromised the ADC_CP and promoted intestinal nitrogen loss via mucus secretion [71]. Notably, the quadratic curvilinear increase of NUA_N with the increase of dietary GKM inclusion levels and unaltered dissolved inorganic nitrogen concentrations (NH₄⁺-N, NO₂⁻-N, NO₃⁻-N) across treatments at the final of the feeding trial, suggested a potential increase in total suspended solids (TSS) nitrogen emissions under high GKM inclusion. Wang et al. reported that partial substitution of dietary wheat flour with kelp meal in formulated feeds significantly elevated TSS levels in hybrid snakehead culture systems [14], also suggesting a potential implication for aquaculture waste management of dietary macroalgae inclusion. However, to the best of our knowledge, there is no information on whether dietary supplementation with macroalgae increases TSS content and/or concurrently elevates nitrogen levels within the TSS fraction in LMB farming, which warrants further mechanistic investigation.

The moderate macroalgae inclusion reduced NUA_P in this study was consistent with that reported by Wang et al., in which a decrease in phosphorus excretion load of hybrid snakehead by feeding the kelp meal inclusion diets was found [14]. The quadratic curvilinear response observed in phosphorus utilization (NUA_P decline with increasing dietary GKM levels) coincided with rising TP in water, indicating a threshold effect: moderate substitutions of SF with GKM reduced particulate phosphorus discharge, whereas full substitution of SF by GKM amplified particulate phosphorus release. The precise mechanism by which altered excretion rates or phosphorus concentration in TSS remains unclear. Given the ecological and operational significance of TSS management, particularly in recirculating aquaculture systems [93], systematic evaluation of macroalgae inclusion effects on both dissolved and solid-phase nutrient dynamics is imperative.

5. Conclusions

Following the feeding trial, LMB that fed the diets containing graded levels of GKM supplementation exhibited significantly enhanced the SR. The study also demonstrated that moderate replacement of SF with GKM substantially improved growth performance and did not reduce the feed utilization efficiency of LMB. Notably, partial substitution of SF with GKM (33% or 67% replacing level) effectively mitigated the whole-body lipid accumulation typically induced by high-starch diets. Furthermore, partial substitution of SF with GKM (33% or 67% SF) has not impaired the NRE and ADC of LMB compared to the control group. Environmental impact assessments revealed the ecological advantages of GKM incorporation, particularly in the GK05 and GK10 groups, showing reduced NUA_P without corresponding increases in NUA_N. These collective findings established GKM as a sustainable functional alternative to conventional SF in feed formulation for the LMB. The dual benefits encompass enhanced aquaculture productivity through optimized growth parameters and environmental sustainability via reduced nutrient discharge. This study provides critical insights for developing eco-friendly aquafeed formulations. Nevertheless, it should be noted that substituting SF with the GKM mixture may increase feed costs, and excessive inclusion levels could preclude its probiotic effects in LMB. And future studies should separately investigate the effects of GL and KG on LMB to facilitate mechanistic interpretation in this research.