Key Performance Indicators of Common Carp (Cyprinus carpio L.) Wintering in a Pond and RAS under Different Feeding Schemes

Abstract

Overwintering impacts common carp performance, yet the nature of changes is not known. The aim of the study was to compare the zootechnical and key performance indicators (KPI) of Cyprinus carpio wintering in a pond with no supplementary feeding (MCF), in a Recirculating Aquaculture System (RAS) fed typical (30% of protein and 8% of fat) carp diet (AFC), and in a RAS fed high protein (42%) and fat (12%) diet (ABF). The analysis showed that ABF fish had the highest final body weight and the Fulton’s condition factor, as well as the lowest food conversion rate compared with AFC and MCF fish. Histomorphological assessment revealed that MCF fish had thinner skin layers, a depleted population of mucous cells in skin, an excessive interlamellar mass in the gills, and no supranuclear vacuoles in the intestine compared to fish from RAS. At the molecular level, higher transcript levels of il-1β and il-6 transcripts were found in the gills of MCF than in fish from RAS. The transcript level of the intestinal muc5b was the highest in ABF fish. Relative expression of il-1β and il-6 in gills were presumably the highest due to lamellar fusions in MCF fish. Described KPIs may assist carp production to ensure sustainability and food security in the European Union.

1. Introduction

Aquaculture is a key food production sector globally [1]. In 2018, its total production was 114.5 Mt in live weight, with a total farmgate sale value of USD 263.6 billion. A considerable part of this production was represented by freshwater aquaculture (51.3 Mt, 62.5%), of which 47 Mt (91.5%) consisted of finfish production. The freshwater production of finfish is forecasted to reach 60% of global aquaculture production by 2030, according to the report, State of World Fisheries and Aquaculture (SOFIA) [2]. Among the five most important aquaculture species, four belong to Cyprinidae: the herbivorous grass carp, Ctenopharyngodon idella (5.7 Mt, 10.5%); the omnivorous common carp, Cyprinus carpio (4.2 Mt, 7.7%); and two planktivorous: silver carp, Hypophthalmichthys molitrix (4.8 Mt, 8.8%), and bighead carp, Hypophthalmichthys nobilis (3.1 Mt, 5.8%) [2]. The latest report published by the Food and Agriculture Organisation (FAO) on the top 10 species groups in global aquaculture in 2019 showed that “carps, barbels and other cyprinids” are the main species group in freshwater aquaculture, amounting to nearly 24.8% (29.8 Mt) of 120 Mt of world production. The publication also showed that production of this group has increased 1.5% from the 2018 level and, consequently confirming the upward trend predicted in the report SOFIA [3]. However, as emphasized by Belton et al. [4], this trend was possible mainly due to intensification rather than horizontal expansion, i.e., increase in production per unit land and water.

For decades, the main farmed cyprinid species in Europe has been common carp (hereafter referred to as carp) that was the mainstay, both traditionally and commercially, for fisheries production in “land-locked” central European Union (EU) [5]. Total production of C. carpio in the EU in 2019 was 65,715 tonnes, and the Czech Republic followed by Poland, Hungary and Germany (ranked in order of production size) provided nearly 80% of carp production in the region [6]. In the EU, carp is produced with the conventional Dubisch method, which involves multiple fish transfers into different stage ponds throughout production. A crucial part of carp farming is wintering (October through April) that happens twice, i.e., after the first and second year of production. In this period, fish do not receive feed and are relatively active; hence, their growth and condition parameters undergo gradual deterioration [7,8]. However, with the existing farming model, carp production has continued its declining trend in the EU (by 10% between 2009 and 2018) due to numerous circumstances, such as predation by protected wildlife (cormorants, great egrets and otters), increasing cost, lower subsidies, and consumer preference for carnivorous finfish such as trout or salmon [6,9]. Carp farmers in the EU are seeking opportunities that would increase the consumption of carp throughout the year, not only in the Christmas period, in which sales reach up to 90% of their annual values [10]. Moreover, consumer awareness of animal welfare causes decreased interest in purchasing live fish and, consequently, a gradual increase in carp processing into a more convenient form, i.e., carcasses, slices, sheets and fillets. These circumstances, but also global indicators, such as climate change (e.g., water shortages, eutrophication), force carp farmers to quickly update their business models accordingly. One way in which the economic profitability and sustainability of carp production can be improved is its eco-intensification, as assessed in the GAIN (Green Aquaculture Intensification) project, which includes shortening the conventional production time (from 33 to 19 months) by running part of the rearing process (first wintering) in a closed recirculating aquaculture system (RAS) decoupled with aquaponics (Figure 1).

Eco-intensification, however, requires verification by assessing a set of key performance indicators (KPI), i.e., growth efficiency (weight gain, feed intake), resource utilisation (food conversion ratio), health and welfare issues (survival rate, enteritis), and in market-size fish, quality (fillet yield, taste). For C. carpio, various bioindicators in muscle [11], liver and intestine [12], and skin [13] have been identified. Moreover, several performance indicators such as total growth rate, survival rate, protein efficiency ratio and feed conversion ratio were commonly used for common carp farmed in cages [14] and ponds [12] during production season. However, to the best of our knowledge, the KPIs in common carp overwintering have never been assessed and are currently not available for this species. Therefore, the aims of the study were to compare the histological and molecular KPI of C. carpio wintering in a pond with no supplementary feeding (MCF), in a RAS fed typical (30% of protein and 8% of fat) carp diet (AFC), and in a RAS fed high protein (42%) and fat (12%) diet (ABF), and to assess whether wintering common carp in RAS can be a viable option to shorten culture time and contribute to the eco-intensification of common carp production in Europe.

2. Materials and Methods

2.1. The Experimental Trial and Fish

The ethical committee of the Faculty of Food Sciences and Fisheries, West Pomeranian University of Technology in Szczecin approved the fish trial. “Guidelines for the treatment of animals in behavioural research and teaching,” published in Animal Behaviour, were adhered to during the trial [15].

The experimental setup was designed to tighten the production cycle from the traditional 33 months to 19 months with subsequent welfare improvement. The fish trial was performed at the Fisheries Research Station (FRS), Nowe Czarnowo, Poland (53°120′36″ N 14°270′48″ E) in a RAS system under a plastic tunnel in ambient, i.e., weather-dependent conditions. One week prior to the trial start, 1620 fish (47.18 ± 1.82 g) were obtained from the carp farm in Maliniec and distributed for acclimatisation (n = 270 per tank) in two sets of three tanks (n = 6), with a tank capacity of 2.7 m³ each (Figure 2).

The 203-day trial (October–May 2019/2020) was performed in triplicate (n = 3 diet⁻¹). Fish were automatically fed with two commercial feed blends, with different levels of protein and fat, i.e., typical Agro-Fish Carp (AFC, 30% and 8%, respectively) and high energetic Aller Bona Float (ABF, 42% and 12%, respectively). Fish were fed diets at a rate of 1.5% body weight using automatic feeders. The rate of each diet was calculated based on information from biweekly bulk weighing of fish. During the trial, the temperature of the water in the RAS tanks was between 6.0 (January) and 26.2 °C (May), oxygen saturation between 40% and 82%, and pH value between 5.75 and 6.75. Traditional wintering was performed in the Maliniec carp farm (MCF), (53°42′5″ N 15°21′22″ E) in an approx. 2.5 m deep wintering pond with a total surface area of 3 ha. During that period, the temperature in the wintering pond was between 4 and 8 °C until May, and feeding was not applied.

2.2. Sample Collection

At the end of the wintering trial, fish (n = 6) from each dietary treatment in RAS (n = 2 per tank) were sacrificed using a lethal dose of 2-phenoxyethanol (2 mL L⁻¹), (Sigma-Aldrich, Saint Louis, MO, USA). Simultaneously, fish from the wintering pond (n = 6) were sacrificed with the same procedure. Briefly, fish gills (2nd and 3rd arches), proximal intestine and head kidney samples were collected immediately and secured in DNA/RNA Shield™ (Zymo Research, Irvine, CA, USA) and, until RNA extraction, were stored at −80 °C. Additionally, a piece of proximal intestine bulb (approx. 5 mm), skin (1 × 1 × 0.5 cm) from the mid-dorsal epaxial body and 2nd and 3rd gill arch samples were collected and washed with deionized water and covered with 10% buffered formalin solution in 50 mL glass jars at room temperature for 5 h [16]. To assess the wintering and nutritional effects on carp in RAS, the following zootechnical parameters were calculated: final body weight (FBW) as an average weight of the carp at the end of the wintering period, and feed conversion ratio (FCR), Fulton’s condition factor (K) and survival rate (SR) using the Equations (1)–(3):

FCR = FC × WG⁻¹

(1)

SR = FN × IN⁻¹ × 100

(2)

K = (W × 100) × L⁻³

(3)

where:

• FC—feed consumed (g);
• WG—weight gain (g);
• FN—final number of individuals;
• IN—initial number of individuals;
• W—fish weight (g);
• L—fish length (cm).

The data for MCF on Fulton’s condition factor (n = 20), final body weight (bulk weighting of 100 fish) and survival rate (based on total biomass) were collected as part of the standard Maliniec carp farm screening after wintering, and none of the fish were harmed for the purpose of this study.

2.3. Total RNA Extraction and Synthesis of the cDNA

All preserved samples were homogenised in 750 μL Tri Reagent^® (Zymo Research, Irvine, CA, USA) for 60 s using Minilys^® homogenizer (Bertin Corp., Rockville, MD, USA). Total RNA was extracted using Direct-zol™ RNA MiniPrep kit (Zymo Research, Irvine, CA, USA), with DNase I treatment, to avoid contamination with genomic DNA. The extracted DNA was quantified and quality checked using NanoDrop 2000 (ThermoFisher Scientific, Waltham, MA, USA) and electrophoresis on 2% agarose gel. The 260/280 ratio of all RNA extracts was 1.8–2.1, and signs of RNA degradation were not observed. Transcriptor First Strand cDNA Synthesis Kit (Roche, Basel, Switzerland) and 1 μg of RNA were used for reverse transcription, following the manufacturer’s instructions, using anchored oligo(dT)18 primers.

2.4. Assessment of Gene Expression in Gills, Intestine and Kidney

Real-time PCR was conducted on LightCycler^® 480 II (Roche, Switzerland). LightCycler^® 480 SYBR Green I Master (Roche, Basel, Switzerland), 0.1 μM of each primer and 5 μL of 10× diluted cDNA templates were used for the reaction in the final volume of 20 μL. All reactions were performed under the following thermal profile: initial activation at 95 °C for 5 min, 45 cycles of denaturation at 95 °C for 10 s, annealing at 60 °C for 10 s and extending at 72 °C for 15 s. At the end of each qPCR melting curve analysis (65–95 °C) was conducted to ensure the specificity of amplification. To ensure the absence of genomic DNA contamination, a random RNA sample was tested accordingly. Relative expression of the common carp genes in the 2nd and 3rd gill arches [immunoglobulin M (IgM), interleukin 1 beta (il-1β), interleukin 6 (il-6), interleukin 8 (il-8), tumour necrosis factor alpha (tnf-α), mucin 5b (muc5b), lysozyme C (lysC), superoxide dismutase 1 (sod1), catalase (cat), glutathione peroxidase (gpx) and glutathione S-transferase (gst)], in the proximal intestine [IgM, il-1β, il-6, il-8, tnf-α, muc5b, lysC, heat shock protein 70 (hsp70) and heat shock protein 90 (hsp90)] and in the head kidney [il-6, il-8, il-1β, tnf-α, IgM, muc5b, lysC, inducible nitric oxide synthase (inos), nuclear factor-erythroid 2-related factor 2 (nrf2)] were measured as well as two reference genes: 60S ribosomal protein L8 (rpl8) and 40S ribosomal protein S11 (40sRNA) (

Table 1. Sequences of C. carpio primers used for qPCR analysis.

Gene	Primer Sequence 5′ -> 3′		Tm (°C)	Function	Reference
IgM 1	F	TCGTATTAGCACCCCCAGAG	53.8	First line of host defence against infections	[18]
	R	TCATCAGCAAGCCAAGACACA	52.4
il-1β2	F	CCTGAAGAAGAGGAGGCTGTCA	56.7	Mediator of inflammatory response	[19]
	R	AAGGAGGCCAGTGGCTCTGT	55.9
il-6 3	F	CCGCACATGAAGACAGTGAT	51.8	Stimulating acute phase protein synthesis	[20]
	R	GGGTATATTTGGCTGCAGGA	51.8
il-8 4	F	TGGAGCTCTTCCCTCCAAG	53.2	Attracting and activating neutrophils	[20]
	R	AGGGTGCAGTAGGGTCCAG	55.4
tnf-α5	F	CCTTGGAAGTGACATTTGCTTTT	51.7	Signalling events within cells	[19]
	R	GCTGTCTGCTTCACGCTCAA	53.8
muc5b 6	F	CAGCCCTCTTCCTCTTTCATC	54.4	Ensure normal mucus clearance	[20]
	R	CCACTCATCTTTCCTTTCTCTTC	53.5
hsp707	F	TGAGAACATCAACGAGCCCA	51.8	Protein maturation, re-folding and degradation	[21]
	R	TTGTCAAAGTCCTCCCCACC	53.8
hsp90 8	F	AAAGACCAGGTCGCCCACTC	55.9	Protein maturation, re-folding and degradation	[22]
	R	AGTACTCGTCGATGGGCTCG	55.9
sod1 9	F	TGGTCCACCGTGAGCTTTATT	52.4	Antioxidant enzyme	[23]
	R	GACAACACAAACGGCGGCAT	53.8
gpx 10	F	TGCAACCAGTTCGGACATCA	51.8	Catalyses the reduction of hydrogen peroxide	[24]
	R	GAAGCCATTTCCAGGACGGA	53.8
cat11	F	CTGGAAGTGGAATCCGTTTG	51.8	Maintaining the cellular redox homeostasis	[25]
	R	CGACCTCAGCGAAATAGTTG	51.8
gst 12	F	TACAATACTTTCACGCTTTCCC	51.1	Protect cellular macromolecules	[26]
	R	GGCTCAACACCTCCTTCAC	53.2
rpl813	F	CTCCGTCTTCAAAGCCCATGT	54.4	Ribosomal protein coding	[27]
	R	TCCTTCACGATCCCCTTGATG	54.4
40sRNA14	F	CCGTGGGTGACATCGTTACA	53.8	Ribosomal RNA gene	[28]
	R	TCAGGACATTGAACCTCACTGTCT	55.7

¹Immunoglobulin M; ²interleukin 1 beta; ³interleukin 6; ⁴interleukin 8; ⁵tumour necrosis factor alpha; ⁶mucin 5b; ⁷heat shock protein 70; ⁸heat shock protein 90; ⁹superoxide dismutase 1; ¹⁰catalase; ¹¹glutathione peroxidase; ¹²glutathione S-transferase; ¹³60S ribosomal protein L8; ¹⁴40S ribosomal protein S11.

).

Decimal dilutions (ranging from 0.92 to 1.10) were performed to evaluate and correct the efficiency of qPCR reactions. The 2^−∆∆Ct method was used to calculate the relative gene expression using the GeneEx (MultiD Analyzes, Göteborg, Sweden) software [17].

2.5. Histomorphology of Gills, Skin and Intestine

Fixed in 10% buffered formalin samples of 2nd and 3rd gill arches, skin from the mid-dorsal epaxial body and proximal intestine fragments were dehydrated using alcohol and were saturated in intermediate solutions (benzene, benzene: paraffin) [16]. Next, the samples were embedded in paraffin blocks and then trimmed and serial sectioned (6 ± 1 μm, Rotary Microtome MPS-2, Opta-Tech, Warsaw, Poland). Gill and skin samples were stained with alcian blue and periodic acid–Schiff (AB/PAS, pH 2.5) [29]. Intestine samples were stained with haematoxylin and eosin (H&E) and alcian blue (AB) [30]. All samples were mounted with DPX mounting medium and covered by coverslips. Twelve glass slides (3 fish × 4 slides) for both RAS diets and traditional wintering and all tissue types were randomly selected and examined using an Eclipse E600 microscope (Nikon, Tokyo, Japan) with 100× objective and the NIS-Elements Basic Research software (Nikon Instruments Europe B.V, Amsterdam, The Netherlands). Gill samples were examined to assess the thickness of the epithelial tissue, and the number and area of the mucous cells (MC). In the skin samples, the thickness of the epidermis was assessed, and the ratio between the thickness of the epidermis (E) and the outer part (stratum spongiosum, SS) and the deeper part (stratum compactum, SC) of the dermis was calculated to compare the share of each layer between ABF, AFC and MCF fish. Moreover, the area of MC and the number of MC per 100 μm of epithelium were counted manually. The total number of acidic, neutral and mixed (acidic and neutral) mucin-containing cells per 100 μm of epidermis was counted. Intestine samples were examined to assess the size of supranuclear vacuoles (SNV), width of lamina propria (LP), number and area of goblet cells (GC) per villus, and thickness of sub-epithelial mucosa (SM). Gill, skin, and intestine samples were also checked against the presence of pathological changes. Measurements of histological section characteristics were made according to the methodology described earlier for different fish species [31,32,33].

2.6. Statistical Analysis and Data Visualisation

Throughout this paper, data are shown as mean ± standard deviation unless otherwise specified. The Shapiro–Wilk test (significance level p < 0.05) was used to assess the normal distribution of data. Depending on the normality of distribution, the ANOVA or the Kruskal–Wallis test and subsequently the Dunnett’s, Tukey HSD and Dunn’s post hoc test were used to assess significance of differences, using Statistica 13.3 software (TIBCO Software Inc., Palo Alto, CA, USA). Visualisation of data was performed using R-package ggplot2 [34] and bio-render.com (access on 6 March 2022).

3. Results

3.1. Basic Performance and Welfare Indices of Carp Wintering in Pond and RAS

Final body weight of MCF fish (56.44 ± 9.80 g) did not differ significantly compared with the initial weight of the fish (47.18 ± 1.82 g). The final body weight of fish from RAS differed significantly (p = 0.01) between AFC (91.01 ± 3.05 g) and ABF (113.30 ± 4.71 g). Fulton’s condition factor K was significantly (p = 0.03) lower for MCF fish (1.81 ± 0.20) compared with fish before overwintering (1.92 ± 0.12). Additionally, K was significantly higher (p = 0.003) for AFC (2.11 ± 0.17) and ABF (2.20 ± 0.19) fish compared with MCF fish and those before experimental RAS. The survival rate did not differ significantly between AFC (99.91 ± 0.19%) and ABF (99.54 ± 0.47%) fish but was higher than in traditional pond wintering (survival rate 70%, pers. comm. M. Gzyl). The feed intake was significantly higher (p = 0.05) for ABF (129.91 ± 3.04 g) than for AFC (120.54 ± 4.61 g), and consequently FCR was significantly lower (p = 0.01) for ABF (1.98 ± 0.09) than for AFC (2.72 ± 0.07).

3.2. Histomorphology of the Gills, Skin and Intestine of Carp Wintering in Pond and RAS

Gill histology of carp demonstrated that the thickness of the epithelial tissue between the secondary lamellae in MCF was lower than in AFC and ABF (p = 0.01), as the ratio between the length of the secondary lamellae and the interlamellar cell thickness was 1.43, 3.96 and 2.92, respectively (Figure 3A–C).

In some cases, in the gills of carp wintering in ponds, extensive epithelial cell hyperplasia led to complete lamellar fusions (Figure 4A). In carp wintering in the RAS, the gills of AFC carp had thicker secondary lamellae than in ABF carp due to epithelial cell hyperplasia (Figure 4B,C) and an increased number of mucous cells (Figure 4D). The number of mucous cells (MC) was significantly higher (p = 0.01) in the MCF group (n = 54 ± 3) than in the AFC (n = 17 ± 3) and ABF groups (n = 10 ± 2). Moreover, the size of MC was the largest (p = 0.01) in MCF (112.6 ± 26.96 μm²) compared with AFC (44.8 ± 9.37 μm²) and ABF (59.1 ± 28.01 μm²) fish. Sections of skin were compared between fish wintering in RAS (ABF, AFC) and ponds (MCF) (Figure 5). The epidermis was thicker in AFC fish (64.71 ± 3.89 μm) than the other two groups, which were also different from each other (ABF 53.39 ± 3.32 μm, MCF 48.47 ± 3.05 μm). The measurements of the epidermis, stratum spongiosum and stratum compactum, and subsequent calculation of the ratio between the thickness of each layer (ABF 1:3.4:4.6, AFC 1:1.2:2.9, MCF 2.3:1:4) showed that, in all fish, SC was the thickest part in the ABF, AFC and MCF fish. The SS was the thickest layer both in ABF and AFC, except the MCF fish, in whom thickness of SS was the least according to the calculated ratio (Figure 5A–C). The overall skin morphology showed higher amounts of subcutaneous fat (adipose tissue) deposits in ABF fish (Figure 5B) than in AFC (Figure 5C). Lack of subcutaneous adipose tissue was observed between SS and muscle tissue in the skin of MCF fish (Figure 5A). The area of MC was the largest in ABF fish (153.5 ± 25.01), (Figure 5D) than in AFC (84.96 ± 19.43) and MCF (46.18 ± 21.43) carp (p < 0.001). The total number of acidic, neutral and mixed (acidic and neutral) mucin-containing cells per 100 μm of epidermis was the lowest in MCF fish (2.8 ± 1.1), (Figure 5E) and between fish from RAS fed typical carp (AFC 4 ± 0.7) and high (ABF 7.2 ± 0.8) fat and protein diets (p < 0.01), (Figure 5F).

Results from the histological evaluation of the intestine samples showed that AFC fish had a significantly (p = 0.01) larger size of supranuclear vacuoles, smaller size of lamina propria, an intermediate number of goblet cells and lower thickness of submucosa compared with ABF and MCF. The GC area was significantly (p = 0.01) larger among AFC and MCF fish compared with ABF fish (Figure 6,

Table 2. Intestine histological parameters of common carp wintering in ponds (MCF) and in RAS fed with typical carp (AFC) and high (ABF) fat and protein diets.

Parameter/Fish Group	MCF	AFC	ABF
SNV (µm)	ND	23.2 ± 3.13 a	18.8 ± 1.67 b
LP (µm)	42.4 ± 11.69 a	19.3 ± 3.93 b	24.5 ± 5.16 c
GC (n)	28.0 ± 2.45 a	57.6 ± 3.44 b	73.5 ± 7.77 c
SM (µm)	72.9 ± 9.68 a	45.1 ± 8.31 b	58.9 ± 4.56 c
GC (µm2)	50.7 ± 13.54 a	50.7 ± 9.95 a	41.7 ± 10.63 b

Abbreviations: SNV, supranuclear vacuoles; LP, lamina propria; GC, number of goblet cell per villus; SM, sub-epithelial mucosa; ND, not detected. Different lowercase letters in rows indicate significant differences between values (p ≤ 0.01). Values were compared with parametric tests.

).

3.3. Gene Expression in the Gills and Intestine of Carp Wintering in Pond and RAS

Gene expression in the gills showed that fish from traditional wintering (MCF) in the pond had a significantly higher relative mRNA expression of il-6 and a high level (over 13-fold change) of il-1ß compared with fish from RAS (Figure 7). In contrast, the IgM and gpx activities were significantly lower in MCF compared with AFC and ABF. The expression of sod1 was similar between both RAS groups and significantly higher in ABF compared with MCF. A difference between AFC and ABF was found in the expression of other oxidative stress-related genes (cat and gst). Expression in MCF was similar to that in AFC and ABF for cat and similar to that in AFC for gst.

The effects of wintering in the pond and RAS on the expression of genes in the intestine of carp are shown in Figure 8. Gene expression analysis in the intestine showed that fish overwintering traditionally in the pond had a significantly lower relative mRNA expression of IgM, il-6, il-8 and hsp90 than fish wintering in the experimental RAS (AFC, ABF). Additionally, the level of il-6 transcripts in the intestine of fish fed AFC diet was significantly lower than in ABF carp. The transcript level of the mucin gene muc5b was the highest in the intestine of ABF fish compared with MCF and AFC carp. Gene expression of the third interleukin il-1ß and the chaperon hsp70 were similar in the intestine of ABF and MCF fish, and significantly higher compared with AFC fish. No difference in the expression of tnf-α between fish wintering in the pond and RAS and between fish from RAS fed AFC and ABF diets were observed.

Analysis of relative mRNA gene expression in the kidney samples showed higher expression of IgM in the ABF than AFC and MCF carp. The transcript level of il-1β was the highest in the MCF fish to the other two groups, which were not different from each other. The expression of the tnf-α and nrf2 genes was the lowest in AFC carp compared to ABF and MCF, which had similar level of transcripts in the kidney samples (Figure 9).

The study also compared relative mRNA expression of lysC, IgM, il-1β, il-6, il-8, tnf-α and muc5b between the samples of gills, intestine and kidney individually for the ABF, AFC and MCF carps. In most of the instances, analysed genes in MCF fish had the lowest expression in the intestine (il-6, il-8, tnf-α) or was similar either to the expression level in gills (IgM) or kidney (lysC). Only in the case of the il-1β there was an intermediate level of expression in the intestine found compared to the two other samples (Figure 10A). In the AFC, all the genes had the lowest expression in the intestine samples for the gills and kidney, which were not different from each other (Figure 10B). Fish fed a high fat and protein diet (ABF) had the lowest transcript levels of lysC, IgM, il-8 and tnf-α in intestine samples. The relative mRNA expression of il-1β and il-6 was higher in gills and intestine than kidney, while the expression of muc5b was significantly lower in kidney than gills (Figure 10C).

4. Discussion

4.1. Differences in Weight Gain, Condition Factor and Mortality of Carp Overwintering in Earthen Ponds and RAS

Multi-stage farming of common carp in Europe, here referred to as the Dubisch style, since the early beginning, has included two wintering periods in >2 m deep earthen ponds until spring when the temperature profile and primary production in ongrowing ponds are favourable to continue production. During wintering that lasts usually from late October to May fish were restricted from feeding and their activity, including metabolic, depending on water temperature [7]. Our study showed minor weight loss and lowered (5.7%) condition in the MCF fish after wintering. According to Lukowicz and Gerstner [35], overwintering of common carp is considered successful if condition factor K does not decrease more than 15–20%. Additionally, Geldhauser and Gerstner [36] showed that weight loss of 5–10% is a typical observation for conditions of Central European aquaculture in winter. The fish from RAS (AFC, ABF) had significantly higher weight and better conditions than those wintering in the pond (MCF), which may be due to more favourable thermal conditions and adapted feeding programme. The differences in the final weight between fish wintering in RAS could be attributed to the composition of the feeds, mainly the content of the lipids. In ectothermic common carp the energy is primarily stored in the form of lipids that must be assimilated before the winter period when animals are decoupled from the resource base [37]. In RAS, the ABF carp were fed with the diet containing a higher content of lipids (12%) than AFC (8%), therefore efficiently utilising this compound to meet metabolic demands and continuing to build body mass.

Fish overwintering in earthen ponds may face stress from starvation, cold, and predators which together or individually may disturb physiology and lead to huge economic losses in aquaculture due to elevated mortality rates [38]. Common carp farming includes two wintering stages (W1 and W2) in the colder months (October through April) in subsequent years. Survival rate of fish after W1 and W2, even concerning severity of winter conditions, is significantly lower in one-summer-old carp (50%) than two-summer-old carp (80–95%) [39]. In our study, carp after W1 reached a much higher survival rate of 70% that presumably stems from the efficient accumulation of lipid reserves during warmer, more productive months and low impact of common stressors such as too high stocking density, low water quality and injuries caused by handling, parasites, and diseases [37,40]. The higher survival rate of MCF fish, as well as insignificant weight loss and decreased condition factor, surprise even more, as small-bodied fish have a smaller capacity for energy storage and a higher mass-specific metabolic rate [41]. Thus, they are more susceptible to over-winter starvation and mortality than larger bodied individuals, i.e., two-summer-old carp which can accumulate larger absolute and relative lipid levels, and in the case of the Maliniec carp farm, experience 10% mortality during W2 ([42], pers. comm. M. Gzyl). Survival rate of AFC (99.91 ± 0.19%) and ABF (99.54 ± 0.47%) carp wintering in RAS was significantly higher than in the fish from earthen pond but was also better in relation to C. carpio cultured in experimental RAS (93.33 ± 5.70) and comparable to the two biofloc technologies (BFT1 97.57 ± 3.70, BTF2 96.67 ± 5.70) and the system with 50% water exchange method (98.67 ± 1.20) [43]. Such low mortality rate of AFC and ABF fish might be contributed to high quality of water, lack of pathogens and feeding that directly influenced the growth and health state of carp in RAS.

4.2. Influence of Wintering Practice on the Histological and Molecular Indices of Carp Gills

The relative mRNA expression of two pro-inflammatory cytokines, namely il-1β and il-6, was upregulated in MCF, while the relative mRNA expression of tnf-α remained comparable to that in AFC and ABF. A high number of il-1β transcripts and, consequently, il-6 may suggest an ongoing inflammation in MCF fish [44]. Moreover, the epithelial cell hyperplasia visible in the histological picture of MCF gills has been previously linked with inflammation processes caused by different pathogens in Atlantic salmon, Salmo salar [45] and rainbow trout, Oncorhynchus mykiss [46]. However, no difference in the expression of il-8 suggests the absence of acute inflammation in favour of chronic inflammation [47]. It is also plausible that changes in cytokine expression are a result of prolonged starvation [48], which in turn creates chronic stress that also influences the immune response of common carp [49]. Moreover, the increased number of MC, and their size, in the gills of MCF fish compared with AFC and ABF suggests more stressful conditions (e.g., water quality) in the wintering pond [50], despite the lack of differences in the transcript level of muc5b, encoding a major gel-forming mucin in mucus.

The upregulation of IgM in ABF and AFC compared with MCF found in our study may be a result of a higher bacteria load (especially of those pathogenic for carp) present in RAS [51,52]. Such conditions result from higher water temperature in RAS compared with the wintering pond, since both bacterial activity and fish productivity depend on water temperature [53]. In contrast, similar muc5b levels in MCF, AFC and ABF may indicate the absence of pathogenic bacteria in both wintering systems, since in channel catfish (Ictalurus punctatus), microbial infection has been seen to upregulate the activity of muc5b in both the gills and the intestine [54]. To comprehensively understand the influence of wintering on carp, high throughput methods (e.g., RNA-seq) combined with histological observations should be implemented.

We revealed differences in the expression levels of oxidative stress response-related genes (sod1, cat, gpx and gst). The higher expression of gst in MCF compared with that in ABF could be explained by prolonged starvation of fish in the wintering pond [55]. Moreover, the significant difference between AFC and ABF, and no difference between AFC and MCF, could be related to the lower energy content in the AFC feed, which presumably did not fully cover the needs of fish during the wintering period [56]. The lower expression of sod1 and gpx in MCF compared with that in ABF suggests a more complex response to winter and different winter-related conditions such as starvation and multiple factors influencing the expression level. For instance, differences in the number of sod1, gpx and cat transcripts could be related to the level of minerals in formulated feeds [57,58]. Additionally, the presence of Zn can inhibit the GST activity in vitro [59], which we observed in vivo at a molecular level, since ABF feed contained the highest Zn level (50 mg kg⁻¹). Water flow could be another factor influencing the downregulation of gpx in MCF, since there is no water flow in the wintering pond, while in RAS, water flows constantly (1.7 m³ h⁻¹), thus creating relatively stable conditions in terms of oxygen level [60]. The lack of water flow in the wintering pond may result in hypoxic and high ammonia conditions [61], which also alters oxidative stress response in fish [62]. Suboptimal conditions in the wintering pond could potentially decrease the expression of sod1 and gpx and cause pathological changes in the gills (complete lamellar fusions), as reported for largemouth bass (Micropterus salmoides) exposed to hypoxia and high ammonia levels [63]. However, this does not explain the higher expression of cat in AFC compared with that in ABF, and the difference is most probably caused by dietary differences, i.e., lipid and protein levels [64,65].

4.3. Influence of Wintering Practice on the Histological and Molecular Indices of Carp Intestine

A lower number of IgM transcripts in the intestine of MCF fish than those from RAS might result from low water temperature (4–8 °C) and ceased feeding during the wintering period. The relatively stable conditions in deep earthen ponds and fish thriving in suboptimal conditions [42] shaped a relatively constant diversity and richness of the gut microbiota [66]. Some of the gut residents are commensal microbiota populations that have been described as recognizable by immunoglobulins [67], thus having a beneficial effect on the immune system of fish [68]. According to Bisht et al. [69], the intestinal bacterial count, including those that maintain gut health, in C. carpio was higher in the winter compared with the summer season. In contrast, the increased expression of IgM in the intestine of AFC and ABF fish wintering in the experimental RAS was enriched by more complex internal and environmental stimulants (microbiota, feeds, water temperature up to 20 °C). As evidenced by Eichmiller et al. [70], differences in the microbiota of common carp may be due to a combination of the effects of diet, habitat usage, temperature and physiology, and the populations of gut residents may adapt via a specific and dynamic interplay with immunoglobulins [67].

Analysis of gene expression of pro-inflammatory cytokines to some extent showed a relatively consistent picture and explained observations identified in the MCF, AFC, and ABF intestine slides during histomorphological evaluation. The relative mRNA expression of il-1β in the intestine of fish from the wintering pond and from RAS fed the ABF diet was the highest and coincided with the highest severity of enteritis found in these fish compared with AFC carp. However, the source of these observations seems to be different, as MCF fish had enteritis due to prolonged starvation during overwintering in earthen ponds. ABF fish were fed diets with a high amount of energy coming from the increased portion of fat and protein in the administered diet. In the case of fish from the pond, seasonal wintering of fish is a typical procedure in carp farming in Central and Eastern Europe [8], and the species is habituated to tolerate winter temperatures close to 0 °C [71]. Liang et al. [72] reported that adaptation of C. carpio to low temperatures is the result of long-term evolution, but differences in the mechanisms of survival at cold temperature have been found among various populations or subspecies. In our study, the high transcript level of the il-1β gene in the intestine of MCF fish is directly related to the regulation of the inflammatory processes that developed due to specific conditions during overwintering. As the inflammatory response of the immune system progresses, il-1β induces expression of subsequent pro-inflammatory genes, such as il-6, il-8 and tnf-α [73]. Apparently, in the case of MCF fish, the severity of physiological changes in the intestine remained at the “typical-for-wintering” level, and further exacerbation of the immune response was unjustified. In contrast, the limited stimulatory relative mRNA expression of the il-1β gene in MCF was not observed in RAS fish, as the expression of both il-6 and il-8 was ramped up in the latter, specifically in ABF carp fed a high fat and protein diet. The cause of the observed differences in the transcript levels of il-1β, il-6 and il-8 between AFC and ABF fish arose directly from the feed composition, as fish in the RAS were managed according to the same procedures. ABF fish fed a more energetic diet reached higher final body weight (113.30 ± 4.71 g) compared with AFC carp (91.01 ± 3.05 g). Furthermore, histological evaluation showed symptoms of diet-induced enteritis (thickened lamina propria and sub-epithelial mucosa, altered supranuclear vacuolization, increased number of goblet cells in the epithelium) in the intestine of ABF carp. Despite the presence of “non-infectious sub-acute enteritis”, as named by Baeverfjord and Krogdahl [74], Fulton’s body condition factor K as well as survival rate did not differ significantly between AFC and ABF fish. These results appear promising, as the RAS fish, heavier and in better condition compared with those wintering in the pond, will be transferred to ongrowing earthen ponds in the spring to reach market size (~1.2 kg) within the following 6–7 months (until November) to be harvested before Christmas. The lack of difference in the relative mRNA expression of the tnf-α gene between fish groups may suggest that in the intestine, neither acute infection nor ongoing pathogenesis of several chronic diseases were present. The plausible explanation for the comparable number of tnf-α transcripts among variants is that this cytokine plays a key role in regulating inflammation, but some of its functions may overlap with il-1β [73].

Significant differences, during both the molecular and histomorphological assessments, were noticed regarding mucin production in the intestine of carp. At the molecular level, the transcript level of the muc5b gene was the highest in ABF fish, and the histological assessment showed apparently large amounts of mucin found in the intestine of these fish. However, in addition to muc5b, the production of mucins is dependent on the expression level of muc2b, and according to Marel et al. [75], the latter gene is the main one expressed in the first and second intestinal segments of C. carpio. Therefore, the massive amounts of mucins in the intestine of ABF fish could also be due to the high activity of muc2b. Nevertheless, our findings show that muc5b transcripts were, in addition to the brain, liver, skin and gills [75], detected in the intestine of common carp. The expression of muc5b in the gills was significantly higher than in the intestine of AFC (p = 0.04) and MCF (p = 0.04) carp, but not in ABF. As the relative mRNA expression of muc5b in the intestine of MCF and AFC fish was similar, the upregulation of this gene in ABF fish was stimulated by the feed. Similar results were obtained by Smith et al. [76] who showed that feeds with different inclusion levels of laminarin, a seaweed-derived ß-D-glucan, led to a significant increase in muc2 and muc4 expression in the intestine. Further studies in carp should also include the assessment of other mucin genes, as the abundance per mucosal fold of the goblet cells was the highest in the intestine of MCF carp, and surface area was the largest in MCF and AFC fish, while the number of muc5b was similarly the lowest.

Tissue damage and the release of danger signals, such as heat shock proteins, may trigger the expression of pro-inflammatory cytokines [77]. According to this, in our study, the high number of hsp70 transcripts in the intestine of ABF and MCF fish may have influenced the expression of the il-1β gene, which subsequently induced the expression of subsequent pro-inflammatory genes, as shown for il-6 and il-8. This mechanism was also described by Li et al. [78] in the various tissues, including the intestine, of common carp after exposure to glyphosate-based herbicide (GBH). Their results also showed that, in addition to hsp70, the number of hsp90 transcripts also increased in response to GBH. In the case of our study, the expression profile of hsp90 differed from that of hsp70, indicating that conditions in the RAS induced higher expression of hsp90 in the intestine of C. carpio compared with fish wintering in the pond. The hsp70 and hsp90 genes collaborate in numerous cellular remodelling reactions, but according to Genest et al. [79], each of them carries out some chaperone activities independently.

4.4. Influence of Wintering Practice on the Skin Histology of Carp from Pond and RAS

Lower thickness of the epithelium of the MCF fish than those from RAS resulted from an adaptive response to wintering conditions, allowing this species to withstand 203-day food limitation and lower temperatures. Similarly, Caruso et al. [80] in European eel (Anguilla anguilla L.) and Somejo et al. [81] in Nile tilapia found that the average thickness of the epidermis was lower in starved fish than those fed. Additionally, Landeira-Dabarca et al. [82] showed that Atlantic salmon deprived from food for 18 days significantly reduced epithelial tissue turnover and activity. This is in line with our other observations showing that AFC fish fed a less energetic diet had significantly thinner epithelium than ABF due to presumably reduced metabolic rate. The thickness of fish skin is also determined by factors such as species, age, and body region [83]. Therefore, to assess the overall influence of the wintering method on the epidermis, SS and SC of the MCF, AFC and ABF carp, the ratio between thickness of each layer was calculated. Here, the lowest thickness of the SS in the MCF carp showed that overall conditions in the wintering directly reduced activity of this layer. Among all the layers, SS is composed of loosely arranged connective tissue and holds fine blood capillaries, nerves, pigment cells and in some species or fish lines, scales [83]. The lower activity and thus the reduced thickness of the SS and epidermis may be restored or even increased after exposure to stressors as shown in the studies with common carp and rainbow trout [84,85]. Feeding ABF carp with a diet containing higher levels of fat and protein increased the area of subcutaneous adipose tissue in those fish. Wang et al. [86] demonstrated that feeding the Nile tilapia with high-fat diets increased the deposits of visceral and subcutaneous fat, whereas starved fish increased ß-oxidation of monounsaturated fatty acids in the subcutaneous layer to cover current metabolic demands and reduced the content of adipose tissue in general. The observation is also in line with the results of image analysis that showed a lower fat density in the subcutaneous region in the starved gilthead seabream (Sparus aurata) [87].

Limited food availability in the wintering pond presumably reduced both the number and the size of the MC in the skin of the MCF fish. Mucous cell population in the skin shrinks in consequence of limited availability of dietary carbohydrates that are essential to synthetise O-glycosylated glycoproteins, i.e., mucins [82]. During warmer months (May through October), unlike during the winter period, carp utilise a combination of natural food and supplementary feeding with carbohydrate-rich (up to 35–45% of the total diet) cereal grains such as wheat, triticale, maize, barley and rye [88]. Kideys and Hartnoll [89] described a decline in mucus layer production as an energy-saving mechanism under adverse environmental conditions. In RAS, carp were fed with diets containing different, i.e., 26% (AFC) and 28.4% (ABF), levels of carbohydrates. The difference in the number and size of MC between AFC and ABF carp only to some extent may have resulted from the level of carbohydrate in the feeds itself. Presumably, the ABF feed with higher digestible energy (16.8 MJ kg⁻¹) than AFC (15.29 MJ kg⁻¹) more efficiently covered nutritional requirements of ABF carp and thus had a positive impact on the population of skin MC. In a study on Atlantic salmon fed feeds based on marine- or plant-derived ingredients and characterised by different energy digestibility, two ways by which plant-based diets can alter volumetric density and size of MC were identified [90]. Reduced production and secretion of skin mucins may expose fish to a higher risk from pathogenic microorganisms, viruses or parasites [91]. In our study, no clinical signs of disease caused by ectoparasites were observed; however, in our opinion, approx. 30% mortality of common carp during overwintering to some extent might have been caused by pathogens such as a ciliated protozoan, Ichthyophthirius multifilis [92], whereas in RAS, the health and welfare of fish were checked during routine daily maintenance, and almost no common carp mortalities were observed in the feeding groups. Losses of the ABF and AFC fish were minimised as a consequence of a well-adjusted feeding programme, and it is difficult to point out the cause.

4.5. Gene Expression in Kidney under Carp Wintering in Ponds and RAS

The increased expression of il-1β in the MCF carp is a commonly observed case of immunometabolism, an interplay between metabolic and immunological processes that was induced by prolonged chronic stress resulting from restricted feeding [93]. According to Liao et al. [94] and Reuter et al. [95] restricted feeding activates the antioxidant system, and once it is insufficient to clear the damage caused by starvation, numerous transcription factors can be activated, such as nuclear factor kappa light chain enhancer of activated B cells (NF-κB). In line with the current biological understanding, NF-κB can induce expression of the il-1β gene, leading to induction of inflammation and subsequent expression of the other pro-inflammatory gene tnf-α [96,97]. However, upregulation of the tnf-α in the kidney of the ABF fish wintering in RAS was presumably triggered by other or additional mechanisms than by the il-1β itself. The plausible explanation for the observed difference between the expression level on the tnf-α in the kidney of the ABF and AFC carp is the composition of the diets. Tumour necrosis factor-α emerged as a key cytokine that influences intermediary metabolism, and in the study on Swiss mice fed a high-carbohydrate and high-fat diet, only the former was able to increase TNF-α concentration in the liver [98]. Diets rich in various carbohydrates were also found to acutely activate NF-κB and subsequently the synthesis of pro-inflammatory cytokines, namely IL-1, IL-6 and TNF-alpha [99]. In the study on blunt snout bream (Megalobrama amblycephala), hepatic NF-κB, TNF-α, IL-1β and IL-6 expression in fish fed diets were all significantly increased, with increasing dietary carbohydrate levels [100]. Our study showed that expression of nrf2, similarly to il-1β, was higher in MCF and ABF fish compared to AFC. However, upregulation of the nrf2 in the MCF and ABF may have different backgrounds. In MSC fish, overwintering in pond with limited access to food decreased the energy state of an organism and plasma glucose levels, leading to activation of the AMPK/SIRT1 axis that enhanced the expression of the nrf2, an oxidative stress regulator [101,102], whereas feeding ABF fish with a diet containing higher content of carbohydrates than AFC could cause physiological stresses and increased renal transcription of the nrf2 in the kidney [103].

A carbohydrate-rich ABF diet possibly also increased synthesis of IgM transcripts in the kidney of carp since NF-κB controls the expression of numerous downstream genes that control cell proliferation, survival, stress responses, and immune functions [104]. Furthermore, a study on grass carp head kidney leukocytes showed that the TNF-α-mediated NF-κB pathway is an important signalling involved in protective immune and inflammatory responses [105]. Higher expression of IgM in the kidney of the ABF carp was exclusively impacted by feed composition, as both AFC from RAS and MCF overwintering in the pond showed similar expression of the gene. Krogdahl et al. [106] showed that feeding Atlantic salmon with diets containing soybean molasses, an alcohol extract of soybean meal, causes an inflammatory response in intestinal mucosa as evidenced by the increased level of IgM. Other studies showed that the difference in the IgM expression in the intestine of gilthead seabream fed a vegetable and fish oil diet further increased when animals were exposed to adverse conditions, e.g., infection with the Enteromyxum leei [107]. However, the underlying mechanisms involved in IgM activity, especially in the kidney of common carp under different wintering conditions, still remain obscure and warrant further studies.

4.6. Differences in Gene Expression between Gill, Intestine and Kidney in Fish under Different Wintering Conditions

Differences in gene-relative mRNA expression emerge directly from the type of tissue that expresses inherent functions and are additionally challenged by internal and external stimuli. Regardless of the fish group (wintering conditions) the lowest gene expression of the seven analysed genes in most cases was observed in intestine samples. It was especially evident in the samples of AFC fish, whereas in MCF and ABF, the differences were more diverse. All the genes are involved in an array of defence mechanisms, and as evidenced in our study, the gills and the kidney are the primary site for their expression. Gills have a large epithelial surface and are constantly exposed to external milieu; thus, its function goes beyond respiratory gas exchange. To respond to all these challenges actively and accurately, the gills are equipped with their own immune system, called the gill-associated lymphoid tissue (GIALT), thereby substantially contributing to overall fish health and survival [108]. In the case of the kidney, a primary lymphoid organ, elevated expression of the genes in the ABF, AFC and MCF C. carpio is a common observation that results from the nature of this organ. Wu et al. [109] showed that comparing other studied organs, including intestine, gene expression of il-1β in the kidney and gills was significantly higher, and the observed difference has dramatically increased in the large yellow croaker (Larimichthys crocea) upon bacterial infection. Our results were also consistent with some other previous studies on Atlantic halibut, Hippoglossus hippoglossus [110], rainbow trout [111] and sea bass, Dicentrarchus labrax [112], where the primary role of gills in immunological responses was evidenced. The lower level of expression of most of the genes in the intestine might be related to the concentration and composition of intestinal commensal bacteria [113]. McEntee et al. [114] underlined that homeostasis in the gut depends on the activity of specialised immune cells that are regulated by a cross-talk with symbiotic microbes in the lumen. The overall composition of the microbiota in the intestine of the ABF, AFC and MCF carp is another variable which likely differed due to food/feed quality and availability and the environment (e.g., water temperature) in which the animals were overwintering. For example, the unique influence of multiple and synergic overwintering parameters on gene expression was shown for il-1β and il-6, which had similar patterns of their transcript level in the gill, kidney and intestine samples from MCF and AFC carps, but not from ABF. The underlying molecular mechanisms in common carp wintering in different conditions still require further and more detailed studies to define the role of symbiotic microbes and environmental factors that determine the performance of fish at the beginning of the next ongrowing stage in spring.

5. Conclusions

Despite being an important stage in a common carp farming, little is known regarding the response of fish to wintering conditions. In this communication, for the first time, we have identified that fish wintering in ponds due to limited access to natural food base and cold-water wintering conditions experienced loss of weight and mortality of 30% compared to common carp from RAS that maintained weight gain and in which only a low number of mortalities was observed during the trial. This observation confirmed the hypothesis that fish overwintering in RAS have better conditions to presumably reach market size in the second, instead of the third, year of production. The histomorphological examination confirmed that fish adapted physiological responses to the wintering conditions by: (i) adjusting the thickness of layers (epithelium, stratum spongiosum, subcutaneous adipose) and changing the population of mucous cells in the skin; (ii) thickening of the interlamellar mass (metaplasia) in the gills; (iii) altering the size of, e.g., supranuclear vacuoles and lamina propria in the intestine. At the molecular level, the relative mRNA expression of the muc5b gene was the highest in ABF fish, and the histological assessment showed apparently copious amounts of mucin found in the intestine of these fish, whereas the transcript level of pro-inflammatory cytokines in gills, namely il-1β and il-6, was upregulated and correlated with severe lamellar fusions in MCF fish. Here, we described that prolonged chronic stress resulting from restricted feeding increased expression of il-1β in the kidney of the MCF common carp, while feeding ABF fish with a diet containing a higher content of carbohydrates than AFC caused physiological stresses and increased renal transcription of the tnf-α and nrf2 in the kidney.

Implementation of eco-intensification can potentially bring economic benefits to carp farmers. Wintering in RAS can save resources required for pond maintenance, fuel and fish for stocking, as well as solve problems with piscivorous predators and water shortages occurring due to climate change. However, investment in RAS facilities can be high, especially for large carp farms; therefore, a true economic analysis should be performed site specific to ensure that eco-intensification of C. carpio will bring financial benefits and sustainability to the carp sector.

This study has generated novel findings, insights and ideas which represent an important milestone for the common carp model and paves the way for more focused studies to be performed in the future with regard to the key performance indicators. The histomorphological changes and expression of the genes can be used as biomarkers to further delineate the influence of wintering conditions on fish performance that is crucial to shorten the time to produce market-sized carp, which can help to ensure both sustainability and food security in the EU.