The greater duckweed S. polyrhiza was cultured using organic manures viz., cattle manure mustard oil-cake and poultry dropping (Sharma et al., 2019). These manures are cheap and easily available. The production cost of greater duckweed in this study was around Rs.14.0/kg. The greater duckweed was collected, cleaned, dried, and ground. The meal was stored at 4°C prior to further use. The moisture, crude protein, crude lipid, carbohydrate, and ash contents of the greater duckweed meal were 75.2, 366.5, 76.2, 300.2, and 181.9 g/kg, respectively. Five isoproteic, isolipidic, and isoenergetic experimental feeds were prepared with graded inclusion of S. polyrhiza meal replacing soybean meal as the primary protein source (Table 1). Fish feeds were formulated using the Winfeed 2.8 software package (WinFeed UK Limited, Cambridge, United Kingdom). The control feed (SP0) contained soybean meal as the only primary source of protein, while in the four experimental feeds, greater duckweed was incorporated at the levels of 5%, 10%, 15%, and 20% of total feed at the expense of soybean meal, wheat flour, corn meal, and sunflower oil (to maintain constant crude protein, crude lipid, and gross energy levels) to produce feeds SP5, SP10, SP15, and SP20. The soybean meal was replaced in a graded manner, which resulted in changes in the proportions of amino acids in the feeds. Therefore, some specific amino acids such as histidine, methionine, lysine, and threonine were supplemented to the feeds based on the reported requirements of common carp (NRC, 2011). The inclusion levels of these four amino acids were determined using the Winfeed software to ensure the requirements of the fish were satisfied. All dry feed ingredients were blended for 10 min and mixed with the oil before warm water was added slowly and everything mixed thoroughly. The entire mixture was placed in the hopper of the twin-screw-extruder (Basic Technology Private Limited, Kolkata, India), and feed pellets were formed with extrusion conditions as follows: cutter 134 rpm; feeder 10 rpm; extrusion 190 rpm; extrusion torque 9.22; heater 1 temperature 65°C; heater 2 temperature 70°C; and final mass temperature 75°C. The diameter of the produced pellets was 1 mm. All feeds were stored at 4°C prior to use. A common difficulty in the use of feeds based on plant ingredients is their palatability to the fish (Rodriguez et al., 1996), but this can be mitigated by the extrusion process. Antinutritional factors such as trypsin inhibitor, phytic acid tannins, oxalates etc., are found in greater duckweed (Cruz et al., 2011). However, the preparation of the feeds by the extrusion technique helped to mitigate the impact of antinutritional factors as high temperature and pressure inactivate many of these factors and control enzymatic rancidity of nutrients (Rokey, 2004; Stadtlander et al., 2019). In this study, the preparation of feed using the extrusion technology improved the digestibility of proteins and starches and destroyed the antinutritional factors present in the feed. The amino acid and fatty acid compositions of the feeds were measured and are presented in Supplementary Tables 1, 2.

Common carp were cultured and sampled following the guidelines of the University of Delhi Institutional Animal Ethics Committee (DU/ZOOL/IAEC-R/2015/07). The fish were collected from a local fish farm and acclimated in the aquarium at the University of Delhi for 1 week during which time the fry were fed the soybean-based control feed. The fry (0.473–0.479 g) were then distributed randomly into 15 glass aquaria (50 L each) with 30 fry per aquarium. The fry of common carp was selected to determine the influence of the plant-based diet on the digestive physiology of fish and to understand the suitability of these diets for early life stages. Each aquarium was connected to an external filtration unit (Sera fil bioactive 130, Germany). Water from the fish culture units was constantly filtered through the filtration unit to maintain ammonia levels of the units. The dissolved oxygen level of water was maintained with the help of an aerator. The carp were then fed one of the five different feeds SP0, SP5, SP10, SP15, and SP20 with three replicate aquaria per dietary treatment. The feeds were distributed ad libitum two times daily at 09:00 and 17:00 h and the weight of feed measured before distribution. Excess (uneaten) feed was collected from each aquarium 1 h after feeding, oven drying, and recording weight. Water quality parameters, namely, temperature, pH, dissolved oxygen, and conductivity were monitored in each aquarium using a probe connected to a portable meter (HQ40d Multiparameter, Hach, United States). The ammonia (NH₃) level was estimated using a probe, connected to Orion Versastar (Thermo Scientific, United States). The nitrite (NO₂^–), nitrate (NO₃^–), and phosphate (PO4^3–) contents were analyzed regularly (APHA, 2017). There was no significant difference in water quality parameters among the dietary treatments throughout the culture period. Temperature, pH, and dissolved oxygen ranged from 25.8 to 28.1°C, 6.66 to 7.56, and 6.45 to 7.48 mg/L, respectively, during the experimental period. The range of ammonia, nitrite, nitrate, and phosphate levels were 0.003–0.0530, 0.291–0.906, 0.956–3.84, and 0.013–0.097 mg/L, respectively, in different treatments and conductivity ranged from 636–811 μS/cm.

After 60 days of culture, the feeding experiment was terminated, and fish was sampled. Fish were starved for 24 h before harvesting, and then, all fish were euthanized with tricaine methane sulfonate (MS222, Sigma, United States), and the weight of individual fish was measured. Four fish from each tank were pooled (four fish/replicate) and stored at -80°C for the assay of whole-body proximate, amino acid, and fatty acid compositions. Three tank replicates were used for all assays (three replicates per diet, n = 3). The digestive tracts of two individual fish per aquarium were collected (two fish/replicate, three replicates; 2 × 3 = 6 fish/diet) for the assay of digestive enzyme activities. The hepatopancreas from individual fish was collected (100 mg) and stored in 1 ml of TRIzol reagent (Ambion, Life Technologies, United States) for the gene expression analysis (four fish/treatment; from two aquaria 1 + 1 fish and two fish from the third aquarium). Specific growth rate (SGR) and feed conversion ratio (FCR) of fish were calculated as follows.

SGR (% body weight/day) = (In final body mass - In initial body mass) × 100/duration of experiment (days).

FCR = feed (dry weight) consumed by individual fish during feeding trial/weight gain (wet weight) of individual fish.

Intestinal samples were freeze-dried and homogenized in ice-cold Milli-Q^® water (1:10) to maintain neutral pH of the extracts. Homogenates were centrifuged for 30 min at 10,000 × g at 4°C, and supernatants were collected for the assay of digestive enzyme activities using fluorimetry (Multimode reader, BioTek Synergy H1 Hybrid, United States) using three replicates per dietary treatment. The amylase activity was determined using an assay kit (E33651; Invitrogen, United States) with fluorescence measured at 485 (i.e., excitation) and 520 nm (i.e., emission). The enzyme activity was expressed as mU/mg protein/min. A protease kit (E6638; Invitrogen) was used to measure the total protease activity with fluorescence measured at 485 (i.e., excitation) and 530 nm (i.e., emission). The protease activity was expressed as fluorescence change/unit. The substrate N-benzoyl-L-arginine-methyl-coumarinylamide (Sigma-Aldrich) was used for the estimation of serine proteases trypsin (Ueberschär, 1988). The fluorescence was measured at 380 (i.e., excitation) and 440 nm (i.e., emission). The chymotrypsin was measured using succinyl-Leu-Val-Tyr-4-methyl-coumaryl-7-amide (Sigma-Aldrich) as the substrate (Cao et al., 2000). The fluorescence was measured at 380 (i.e., excitation) and 450 nm (i.e., emission). The enzyme activities were expressed as μM 7-amino-4-methylcoumarin (AMC)/mg protein/min. The neutral lipase activity was determined following the method of Roberts (1985) using the substrate 4-methylumbelliferyl butyrate, 4-MU (Sigma-Aldrich). The fluorescence was measured at 365 nm (excitation) and 450 nm (emission). The enzyme activity was expressed as μM 4-MU/mg protein/min. Protein content was estimated using bovine serum albumin (BSA, Sigma-Aldrich) as standard (Bradford, 1976).

Samples of feed were ground, and the fish samples (four fish pooled per replicate) were blended to form homogeneous paste prior to biochemical composition analyses. Proximate compositions of feeds and cultured fish were determined (three replicates for each assay) following the standard protocols of the Association of Official Analytical Chemists International (AOAC, 2000). The moisture level was calculated after drying samples at 110°C for 24 h, and ash content was measured after incineration of samples at 600°C for 16 h. The nitrogen content was first assayed using micro Kjeldahl method, and then, crude protein was calculated (N × 6.25). The crude lipid content was measured gravimetrically following extraction of total lipid using chloroform/methanol (2:1, v/v) (Folch et al., 1957). The subtraction method was applied for the calculation of carbohydrate levels in the feeds (Aksnes and Opstvedt, 1998).

Feed and whole fish samples were processed as described above and amino acid compositions assayed with an Automatic Amino Acid Analyzer L-8900 (Hitachi Co., Ltd., Tokyo, Japan) using three replicates for each sample. For all amino acids other than cysteine, methionine, and tryptophan, samples were hydrolyzed using 6 N HCl at 110°C for 24 h (Bassler and Buchholz, 1993; Chakrabarti et al., 2018). The sulfur-containing amino acids were analyzed after oxidizing the sample with performic acid prior to treating the sample with 6 N HCl. For the estimation of tryptophan, the sample was hydrolyzed using 4 M methanesulfonic acid with 0.2% 3-(2-aminoethyl). A nitrogen evaporator (PCi Analytic Private Limited, Maharashtra, India) was used to dry the digested samples, and 0.02 N HCl was added to obtain a protein concentration of 0.5 mg/ml in the sample, and 1.5 ml was placed in a glass vial of the auto sampler. A cation-exchange resin column (4.6 mm ID × 60 mm L) with 3 μm particle size was used for the separation of amino acids with the following analytical conditions: column temperature 30-70°C, reaction temperature 135°C, and a ninhydrin flow rate of 0.35 ml/min. All amino acids were monitored at 570 nm, except proline and hydroxyproline that were monitored at 440 nm. The concentration of individual amino acids was compared with a standard solution (Wako Pure Chemical Industries Limited, United States) and expressed as g/kg.

After processing feed and whole fish samples as described above, fatty acid compositions were analyzed by gas chromatography and flame ionization detection (GC-FID) using a Clarus 580 (PerkinElmer, Waltham, United States). In brief, crude lipid was extracted from the samples using chloroform/methanol (2:1, v/v) following the protocol of Folch et al. (1957) with three replicates per dietary treatment. Fatty acid methyl esters (FAME) were prepared from crude lipid extracts by acidic transesterification, treating the lipid with 1% sulfuric acid in methanol for 16 h at 50°C (Christie, 2003). After extraction and purification of FAME (Tocher and Harvie, 1988), a 1 ml aliquot was placed in a glass vial of the GC autosampler. Fatty acids were separated using a 60 m ZB-wax GC column, internal diameter of 0.32 mm, and film thickness of 0.25 μm (Phenomenex, Hyderabad, India). Data were collected using preinstalled programmed software (TotalChrom Workstation Ver6.3; PerkinElmer, United States). The FAME were identified and quantified by the comparison with standards (Supelco FAME 37 mix; Sigma-Aldrich, United States) and published data (Tocher and Harvie, 1988) and concentration expressed as mg/100 g.

The levels of mRNA expression of delta-6 fatty acyl desaturase (fads2d6), elongation of very-long-chain fatty acids protein 2 (elovl2), elongation of very-long-chain fatty acids protein 5 (elovl5), and fatty acid synthase (fas) genes were determined in the hepatopancreas of common carp. Total RNA was extracted using the TRIzol reagent (Ambion, Life Technologies, United States) following the protocol of the manufacturer. The absorbance of extracted RNA was examined at 260 and 280 nm using a Nanodrop spectrophotometer (Thermo Scientific, United States) to determine the concentration and quality. The extracted RNA was treated with 1 U of DNase I (Sigma-Aldrich, United States) to avoid DNA contamination, and the quality of RNA treated with DNase was checked with 1% agarose gel electrophoresis. Subsequently, total RNA was reverse transcribed into cDNA by the reverse transcription reaction using high-capacity cDNA reverse transcription kit (Applied Biosystems, United States), using the protocol provided by the manufacturer.

Quantification of gene expression was carried out by using quantitative reverse transcription polymerase chain reaction (qRT-PCR) using a Quant Studio 6 Flex system (Applied Biosystems) and PowerUp SYBR™ Green Master Mix (Applied Biosystems). Primers were designed using the online primer design tool of NCBI with β−actin used as the reference (housekeeping) gene (Supplementary Table 3). The efficiency of primers was evaluated by using the melt curve and standard curve analysis using the QuantStudio 6 Flex Real-Time PCR system software v1 (Applied Biosystems). The 10 μl reaction mixture for qRT-PCR was composed of 0.25 μl PCR forward primer (2.5 μM), 0.25 μl PCR reverse primer (2.5 μM), 1 μl of cDNA (1:3), 5 μl of 2 × PowerUp™ SYBR™ Green PCR Master Mix (Applied Biosystems), and nuclease-free water (3.5 μl). Samples were run in duplicate for each target gene with non-template control (NTC). The thermal cycling conditions were as follows: predenaturation of nucleic acid at 95°C for 10 min followed by either 40 cycles of 15 s at 95°C and 1 min at 60°C (primer Tm 60°C) or 40 cycles of 15 s at 95°C, 15 s at 55°C, and 1 min at 72°C for (primer Tm < 60°C). The data of qRT−PCR were calculated using the 2^–ΔΔCt (Livak and Schmittgen, 2001) method with β−actin as the internal control.

Data are presented as means with standard error (SEM) with n values as stated. The IBM SPSS 25.0 software (SPSS Inc., Michigan Avenue, Chicago, IL, United States) was used for the statistical analysis. Data were analyzed using one-way analysis of variance, and the Tukey’s test was performed to compare the differences among experimental groups. The linear regression analysis was performed to check the effect of inclusion level of S. polyrhiza in the diets on growth performance, digestive enzyme activities, proximate composition, and amino acid and fatty acid profiles of fish. The broken-line regression analysis was performed for average weight and SGR to determine the breakpoint, BP (Muggeo, 2008) with final BP estimated based on the least sum of squares of deviation (LS method) using R package. The significance was accepted at p < 0.05 level.

After 60 days of culture, the number of common carp in each aquarium was recorded. There was no mortality of fish; all fish survived (Table 2). The broken-line regression analysis showed the impact of greater duckweed on the growth performance and SGR of common carp. The estimated break point for final weight and SGR were 10 and 13.4% of greater duckweed, respectively. The final weight of SP20 diet-fed common carp was significantly higher compared with other diet-fed fish. Minimum weight was found in the SP0 treatment. Similar trend was also found with SGR. SGR was maximum in SP20 diet-fed carp. FCR was minimum and maximum in SP20 and SP0 treatment, respectively.

The amylase activity ranged from 98.91 to 126.48 mU/mg protein/min in five different diet-fed common carp with a minimum amylase activity in SP5 diet-fed fish (Table 2). The amylase activity was significantly higher in SP0, SP15, and SP20 diet-fed fish compared with other diet-fed fish. The total protease activity was significantly higher in SP10, SP15, and SP20 diet-fed fish compared with other diet-fed fish. Trypsin activity was significantly higher in SP20 diet-fed fish compared with other diet-fed fish. Significantly higher chymotrypsin and lipase activities were found in SP5 and SP15 diet-fed common carp, respectively, compared with other treatments. The inclusion of greater duckweed in the carp diet showed linear relationships (R² = 0.917–0.998) with amylase, trypsin, chymotrypsin, and lipase activities.

Expressions of various genes involved in the metabolism of fatty acid were recorded in the hepatopancreas of common carp fed the five diets. The expression of fads2d6 was significantly higher in fish fed diet SP5 compared with others (Figure 1). The expression levels of elovl2 and elovl5 were significantly lower in SP20 diet-fed common carp compared with others (Figures 2, 3). The significantly higher expression of fas was observed in SP5 compared with others (Figure 4). The mRNA expressions showed polynomial 3 order relationships with different treatments.

The effect of dietary inclusion of greater duckweed at four different levels (replacing soybean meal) on the performance of common carp was recorded in this experiment. The survival rate of fish was not affected with the inclusion of greater duckweed in the diets. The earlier study showed mixed results. The inclusion of greater duckweed more than 20% in diet resulted into mortality of tilapia (Fasakin et al., 1999, 2001), supplementation of L. minor (20%) in diets affected the survival rate of common carp (Yılmaz et al., 2004). El-Shafai et al. (2004) reported that the inclusion of duckweed in the feed of tilapia improved the survival rate of fish. The incorporation canola meal at 50% level in the diet was not affecting the survival rate of tilapia (Iqbal et al., 2021).

The broken-line regression showed that incorporation of greater duckweed at 10% level as a breakpoint for final weight of common carp, and it was 13.4% for SGR of fish. Highest growth performance was observed in SP20 diet-fed fish. Food was efficiently utilized in this treatment as minimum FCR was recorded. In L. minor supplemented diet-fed common carp and tilapia, similar trends of growth performance, SGR, and FCR were found (El-Shafai et al., 2004; Yılmaz et al., 2004). The inclusion of L. polyrhiza in the diets of mrigal Cirrhinus mrigala and rohu Lebeo rohita improved the weight gain, SGR, and FCR (Bairagi et al., 2002; Ghosh and Ray, 2014). Inclusion of fermented L. minor at 2.5% level and canola meal at 50% level increased the growth of tilapia (Herawati et al., 2020; Iqbal et al., 2021). In this study, the SGR of common carp (initial weight: 0.473–0.479 g) ranged from 2.01 to 2.93%. Similar results were reported in earlier study like, in L. minor supplemented diet-fed common carp (initial weight: 0.283–0.295 g), SGR ranged from 1.96 to 2.26% (Yılmaz et al., 2004), and in soy protein concentrate (SPC)-incorporated diet-fed common carp (initial weight: 2.43–2.47 g), SGR was 2.01–2.93% (Zhu et al., 2020). Xie et al. (2021) reported that feeding of common carp with diets containing fishmeal and ultra-micro-ground mixed plant proteins (uPP)-based diets resulted in 540–560% growth of fish after 112 days of culture. In this study with common carp, 238–474% weight gain of fish was recorded after 60 days of culture.

The study of digestive enzyme activities in five different diet-fed common carp explained the reason of efficient utilization of consumed diet in SP20. Total protease and trypsin activities were maximum in SP20 diet-fed fish; considerable amylase and lipase activities were also found in SP20 diet-fed common carp. The digestive enzymes, namely, protease, lipase, and amylase played a significant role in digestion and absorption of nutrient (Zhou et al., 2010). Fish fed with different diets are able to adjust the activity of their digestive enzymes (Shiping and Zhao, 2005). The inclusion of duckweed in the diets of rohu and tilapia enhanced digestive enzyme activities like amylase, trypsin, and chymotrypsin (Goswami et al., 2020; Zhao et al., 2020).

The proximate composition study showed that the inclusion (10–20%) of greater duckweed enhanced the crude protein content of common carp in this study. The crude lipid and ash contents of common carp increased in a graded manner with the enhanced inclusion of greater duckweed in the diet. An earlier study showed that supplementation of duckweed improved the crude protein and crude lipid contents in fish (El-Shafai et al., 2004; Yılmaz et al., 2004; Fasakin, 2008; Abou et al., 2011). Aslam et al. (2021) reported significantly higher crude protein contents in L. minor-incorporated diet-fed grass carp Ctenopharyngodon idella and silver carp Hypophthalmichthys molitrix compared with the soybean-supplemented diet-fed fishes. The inclusion of duckweed increased the ash content in fish (Fasakin et al., 1999; El-Shafai et al., 2004; Fasakin, 2008). This indicated that greater duckweed-supplemented diets fulfilled the nutritional requirements of common carp. The proximate composition study showed that the crude protein, crude lipid, and ash contents of greater duckweed were 36.65, 7.62, and 18.19 g/100 g (Sharma et al., 2019). The amino acid profile of greater duckweed is comparable with soybean meal. Feeding with greater duckweed-supplemented diets improved the non-essential amino acid contents in the common carp. The supplementation of fermented L. minor in the diet enhanced the lysine content in tilapia (Herawati et al., 2020).

In this study, fatty acid composition of fish was influenced by the supplementation of greater duckweed. Highest and lowest SFA contents were found in SP5 and SP20 diet-fed common carp, respectively. MUFA content showed an inverse relationship with the increased inclusion of greater duckweed in diet of common carp. A direct relationship was found between the amount of greater duckweed in the diet and n-6 PUFA and n-3 PUFA contents in common carp. Inclusion of greater duckweed in the diet enhanced the ALA, DHA, and EPA contents in common carp. The duckweeds are a rich source of fatty acids (Appenroth et al., 2017; Chakrabarti et al., 2018; Sharma et al., 2019). The feeding of Azolla filiculoides enhanced the total n-3 PUFA (especially EPA and DHA) content (Abou et al., 2011), and fermented L. minor enhanced LOA (Herawati et al., 2020) in Nile tilapia. Similarly, the contents of EPA and DHA in common carp increased linearly with increasing greater duckweed level in the diet.

In this study, the expression levels of key genes involved in the biosynthesis of fatty acids like fads2d6, elovl2, elovl5, and fas were evaluated in the common carp. Upregulation of all these genes was found in fish fed greater duckweed supplemented diets compared with control diet-fed fish. This might be due to the presence of LOA and ALA in the experimental diets. Earlier study showed that the higher contents of LOA and ALA upregulated the expression of desaturases/elongases (Tocher et al., 2004; Turchini et al., 2006; Francis et al., 2007; Li et al., 2008). However, an excess of ALA in diet can inhibit the transcription of fads2d6 gene (Bell et al., 1993). In this study, among the fish fed the experimental diets, the highest expression levels of genes were recorded in the SP5 treatment, and then the expression gradually decreased. The ALA content increased with increasing inclusion of greater duckweed in diet. EPA and DHA contents increased with decreasing expression of genes elovl2, elovl5, and fads2d6. Similar results were found in common carp and rainbow trout Oncorhynchus mykiss where the expressions of desaturases and elongases were higher in fish with lower contents of EPA and DHA (Ren et al., 2012; Lazzarotto et al., 2018).