Age and Feeding Habits of Trematomus bernacchii in the Ross Sea

Abstract

Trematomus bernacchii is a widely distributed, commercially valuable species that plays a significant role in the Benthic ecosystem of the Southern Ocean. Current research mainly explores its molecular responses to cold adaptation and environmental stress, while questions regarding its biological characteristics, such as length-related and sex-related differences in diet, remain to be addressed. This study assessed the age and feeding habits of T. bernacchii in the Ross Sea through stomach content morphological analysis and DNA metabarcoding analysis, examining dietary shifts by growth and sex. The research revealed that T. bernacchii primarily feeds on fish eggs, polychaetes, echinoderms, and mollusks. Other dietary components include algae, cnidarians, and crustaceans. Minor dietary differences were found between sexes. Polychaetes dominated all size classes, with Crustacea and Cnidaria declining as body length increased, while Echinoderms and Mollusca were more abundant in larger specimens. Evidence of egg cannibalism was also noted. These findings enhance our understanding of the trophic ecology of T. bernacchii and contribute to the construction of the Benthic food web of the Ross Sea.

1. Introduction

The emerald notothen, Trematomus bernacchii, commonly known as the emerald rock cod, is widely distributed in the Southern Ocean and is abundant at shallow depths in the Ross Sea [1]. This Benthic fish is a strict stenotherm, inhabiting coastal regions of Antarctica where it is adapted to persist in extremely low and stable temperatures (−1.86 °C) [2]. As a dominant species in its habitat, T. bernacchii primarily feeds on Benthic prey and exerts a significant influence on other organisms through trophic interactions [3,4,5]. It plays a crucial role in the high-Antarctic food web by linking upper and lower trophic levels and maintaining marine ecosystem stability [6]. T. bernacchii demonstrates strong adaptation to cold, making it highly vulnerable to temperature fluctuations. Due to its local prevalence and ease of capture, it has been a key species in studies on the upper limits of thermal tolerance in Antarctic fish [7]. At the same time, T. bernacchii is widely used to explore the impact of temperature changes on gene expression in Antarctic fish tissues, providing valuable insights into cold adaptation mechanisms and adaptive radiation, while highlighting its unique structure and essential role in cold adaptation physiology [8,9]. Additionally, T. bernacchii is an important subject of research on microplastic pollution. It is a valuable bioindicator for detecting environmental contaminants, as it bioaccumulates pollutants through the food chain and has been used to monitor anthropogenic microparticles [10,11]. T. bernacchii is thus well-suited for assessing the ecological impacts of human activities, and the increasing prevalence of synthetic fibers can be interpreted as a reflection of heightened human activities in Antarctica [12]. Despite its ecological and scientific significance, the biology and ecology of T. bernacchii remain incompletely understood, necessitating further research.

Accurate assessments of age and growth are crucial for fisheries science and the effective management of Antarctic fish populations [13]. With ongoing research into the marine biological resources of Antarctica, this work holds both scientific and economic significance [14]. The age determination of Antarctic and Patagonian toothfish in the Ross Sea provides insights into the operational status of Antarctic fisheries [15,16]. In past studies, methods such as direct observation, length–frequency distribution analysis, radioactive element identification, and analysis of calcified tissues have been used for determining the age of fish [17]. The analysis of calcified tissues, including scales, fin rays, vertebral bones, and otoliths, represents the most widely employed method for determining fish age and has become the recognized standard for understanding and assessing fish populations [18,19]. In research on Antarctic fishes, age estimation has been conducted through the examination of annual growth rings on polished cross-sections of sagittal otoliths, along with the assumption that one opaque zone and one hyaline zone correspond to a single year of growth in the otolith [20,21]. Past research has found that the error in age determination through otoliths decreases as the age of the fish increases [14]. The relatively slow growth and long lifespans of Antarctic fish suggest that using otoliths for age determination yields higher accuracy [22].

Fish feeding habits represent a fundamental aspect of understanding ecosystem dynamics, serving as the basis for constructing marine food webs and examining the nutritional flows within ecosystems [23]. Investigating these feeding patterns offers valuable insights into fish biology and population interactions, making it a critical approach in fisheries ecology research [24]. Typically, studies on feeding habits are conducted using the traditional visual inspection method based on the morphological analysis of stomach contents. This method’s continued use is largely attributed to its operational simplicity and intuitive nature [25]. The study of fish diet composition reveals how prey availability and habitat conditions influence their feeding habits and trophic dynamics [26,27,28]. Similarly, the annual variation in food intake of Antarctic ice fish was examined to gain a deeper understanding of the amount of food consumed by Antarctic demersal fish during the summer [29]. This can also be applied to the early life stages of fish, and it revealed that the diet composition of Chionodraco hamatus during its early life stages consisted almost exclusively of notothenioid fish larvae [30]. However, this method relies heavily on extensive labor and a high level of manual identification skills. Additionally, the accuracy of this method is influenced by the physical characteristics of the prey itself and the extent of digestion [31]. The integration of DNA metabarcoding technology represents an advancement in determining feeding habits, offering higher taxonomic resolution and thus enabling more precise results. For example, DNA metabarcoding has been shown to effectively identify highly digested stomach contents, revealing a more accurate revelation of the high seasonality in fish prey contribution and richness in the study of Merluccius merluccius feeding habits [32]. DNA metabarcoding provides higher taxonomic resolution compared to traditional morphological analysis [33]. By combining DNA metabarcoding with morphological analysis, researchers can achieve a more comprehensive assessment of diet [34]. This approach helps overcome the challenges posed by highly digested prey that lack identifiable physical features [35].

Despite significant progress in understanding various marine species, knowledge about Trematomus bernacchii in the Ross Sea remains incomplete. This study seeks to elucidate the biological and ecological traits of T. bernacchii in this region by conducting a detailed analysis of its age and feeding behaviors. Specifically, the primary objective is to characterize the feeding habits of the species, with age determination conducted alongside this main aim. Through these investigations, we aim to enhance our comprehension of its biology and ecology, thereby providing a foundation for further research on the fundamental biology and feeding ecology of T. bernacchii.

2. Materials and Methods

2.1. Sampling

The samples originated from the 38th Chinese National Antarctic Research Expedition, conducted from November 2021 to April 2022. Fish specimens were collected using traps (1 m × 0.5 m × 0.35 m) in the Ross Sea area (74°56.48′ S, 163°43.72′ W) by the vessel Xuelong. Six traps were deployed together, with pork and chicken used as bait. The traps were set between 05:00 and 16:00 on 18 January 2022, at a sampling depth of 120 m. During the operation, the vessel remained stationary, and heavy weights were used to stabilize the traps on the seafloor, preventing any scraping of the ocean floor. A total of 25 fish specimens were collected, which were identified as T. bernacchii after morphological examination and DNA barcoding. Data collected from all individuals included weight (W, g), gutted weight (GW, g), standard length (SL, mm), total length (TL, mm), and anal length (AL, mm) [1]. Additionally, sagittal otoliths were collected, and the gonads of both male and female individuals were dissected and observed.

2.2. Age Determination

The transverse sectioning method proved to be the most effective for age determination, as it enhances the visibility of areas with slow growth [20]. The otoliths were embedded in epoxy resin, which was cured at 50 °C for 24 h. Then, transverse sections (TS) through the core were made using a precision saw equipped with a diamond-edged wafering blade. Ultimately, the sections were ground and polished on both sides to make the core visible. The section typically displays a regular pattern of translucent and opaque areas, starting from the core along the ventral growth axis to the edge. Each combination of an opaque zone and a translucent zone is considered a ring [36], and the age of each fish is determined by counting these rings [4].

In age estimation, employing a method of double reading by two individuals results in more accurate outcomes. Under conditions where the size is unknown, two readers perform independent readings. All readings are conducted without prior knowledge of the biological information of the specimens [4].

2.3. Stomach Content Analysis

2.3.1. Morphological Identification of Stomach Contents

After dissection, the stomach section, from the end of the esophagus to the pyloric sphincter, was carefully excised and weighed. The stomach contents of T. bernacchii were then stored at −80 °C for future analysis. For stomach content analysis, the gastrointestinal samples were first removed and thawed in a 4 °C environment. A scalpel was used to remove the contents from the stomach. To ensure sterile working conditions, all equipment used, including Petri dishes and scalpels, were sterilized by autoclaving. The dissection process was carried out in a laminar flow cabinet to maintain a sterile environment throughout the procedure. Based on morphology, species were identified to the lowest possible taxonomic level using a Jiangnan JSZ5A stereomicroscope and processed accordingly. After drying the surface moisture of the prey types with filter paper, they were weighed on an electronic balance with a precision of up to 0.001 g. The percentage measurements (F%, N%, W%) were used to evaluate the importance of the prey organisms identified in the diet composition [37].

2.3.2. Molecular Identification of Stomach Contents

DNA extraction was initiated by adding 1000 μL of CTAB lysis buffer and lysozyme to a 2.0 mL Eppendorf tube, followed by thawing the sample at 4 °C and mixing the stomach contents. The sample was then incubated in a 65 °C water bath to ensure complete lysis. Afterward, phenol, chloroform, and isoamyl alcohol were added (25:24:1 ratio), mixed, and centrifuged at 12,000 rpm for 10 min. The supernatant was transferred to a new tube, mixed with isopropanol, and incubated at −20 °C to precipitate the DNA. The pellet was washed with 75% ethanol, air-dried, and rehydrated with ddH2O. RNA was removed by incubating with 1 μL of RNase A at 37 °C for 15 min [38].

For PCR, the reactions were set up with a total volume of PCR mix containing 15 μL of Phusion^® High-Fidelity PCR Master Mix (New England Biolabs, Ipswich, MA, USA), 0.2 μM of each primer, and 10 ng of genomic DNA template. The primers mlCOlintF and jgHCO2198 were used to amplify the target gene fragment COI (

Table 1. PCR primers used in gene cloning and expression in the COI region.

Target Region	Primer Names	Primer Sequences
COI	mlCOlintF	GGWACWGGWTGAACWGTWTAYCCYCC
	jgHCO2198	TANACYTCNGGRTGNCCRAARAAYCA

). The amplification protocol began with an initial denaturation at 98 °C for 1 min, followed by 30 cycles of denaturation at 98 °C for 10 s, annealing at 50 °C for 30 s, and extension at 72 °C for 30 s. A final extension was carried out at 72 °C for 5 min [39].

Sequencing libraries were generated, indexed, and quality-checked using Qubit and real-time PCR for quantification, and a bioanalyzer for size distribution. Quantified libraries were pooled and sequenced on Illumina platforms according to the effective library concentration and data requirements. Sequencing was performed using the NovaSeq6000 system with PE250.

In this study, we divided the original off-machine sequencing data into samples based on Barcode sequences and PCR amplification primer sequences. Subsequently, the FLASH software (version 1.2.11) was used to assemble the paired-end reads of each sample to generate the original Tags data (Raw Tags) [40]. After assembly, the Cutadapt software (Version 3.3) was used to trim and remove the reverse primer sequences and other unwanted sequences to ensure they did not interfere with subsequent analysis [41]. Then, the assembled Raw Tags were subjected to quality control using fastp software (version 0.23.1), rigorously filtering to obtain high-quality Tags data (Clean Tags) [42]. Finally, by comparing the Tags sequences with the species annotation database, chimeric sequences were detected and removed to obtain the final effective data (Effective Tags) [43]. Thereafter, the Uparse algorithm (Uparse v7.0.1001) was used to cluster all the Effective Tags from all samples, grouping the sequences into operational taxonomic units (OTUs) with 97% consistency. The algorithm also selected representative sequences for the OTUs, typically the most frequently occurring sequences within the OTUs, for subsequent species annotation. These steps ensured the quality and reliability of the data [43].

2.3.3. Molecular Identification of an Unknown Fish Egg

During the dissection of Trematomus bernacchii, a significant number of fish eggs were found in the stomach of one of the samples (Figure 1). Due to the limitations of morphological analysis, it was not possible to determine the species of the eggs. Consequently, we employed DNA barcoding techniques to obtain genetic information for species identification. To ensure sampling accuracy, we conducted two independent egg samplings to minimize the influence of host genes. The DNA from the sampled eggs was extracted using a plasmid mini kit provided by AXYGEN. A total of 2 g of fish eggs were crushed, and DNA was subsequently extracted from the homogenized material using the AXYGEN plasmid mini kit. The amplified region was the 12S rRNA gene (

Table 2. PCR primers used in gene cloning and expression in the 12S region.

Target Region	Primer Names	Primer Sequences
12S rRNA	MiFish_U-F	GTCGGTAAAACTCGTGCCAGC
	MiFish_U-R	CATAGTGGGGTATCTAATCCCAGTTTG

), a commonly used marker for marine fish species identification [44,45,46]. The sequencing results were compared using SeqMan in the DNAStar software (Lasergene 7.1) package supplemented with manual correction. These sequences were then matched against the NCBI database using Blastn to annotate the corresponding fish species. To construct a phylogenetic tree for the unknown fish eggs, we used the gene sequence of T. bernacchii as the root to anchor the tree and validate the species of the unknown fish eggs. We conducted a phylogenetic analysis using the Maximum Likelihood method in MEGA11 software (Version 11.0). The phylogenetic test employed the Bootstrap method with 1000 replicates. We selected the Tamura–Nei substitution model and assumed uniform rates among sites, including all sites when handling missing data. During tree inference, we used the Nearest-Neighbor-Interchange (NNI) heuristic method. The initial tree was generated automatically, with the default settings using the Neighbor-Joining (NJ) or BioNJ methods.

3. Results

3.1. Size and Age

There were 14 females, 8 males, and 3 individuals whose sex was indistinguishable due to incomplete gonadal development based on all samples. The standard length range for T. bernacchii was between 144 mm and 258 mm, with an average standard length of 206.7 mm. The total length ranged from 168 mm to 298 mm, with an average of 237.1 mm. The anal length ranged from 76 mm to 156 mm, with an average of 117.3 mm. The age distribution ranged from 10 to 19 years. The weight ranged from 65.0 g to 426 g, with an average of 209 g. The gutted weight ranged from 52.2 g to 301.9 g, with an average gutted weight of 163.88 g (

Table 3. Biological measurements of T. bernacchii. “*” indicates that the sex was indistinguishable.

Age	♂	♀	*	Standard Length/mm	Total Length/mm	Anal Length/mm	Weight/g	Gutted Weight/g
10	—	—	1	144	168	76	65	52
12	—	1	1	164–186	188–212	90–93.5	90–146	76–124
13	3	2	1	169–225	194–258	82–123	93–259	80–209
14	3	1	—	191–231	220–263	95–156	135–274	117–229
15	1	4	—	196–234	228–267	96–134	159–269	113–220
16	—	3	—	212–213	243–246	111–116	217–231	164–191
17	—	1	—	227	259	118	298	217
18	—	2	—	237–258	272–298	129–134	271–426	224–301
19	1	—	—	230	263	120	285	209

).

3.2. Feeding Habits

3.2.1. Morphological Identification of Stomach Contents

Based on the classification characteristics of the stomach contents, this study categorized the prey organisms of T. bernacchii into 10 groups, namely, Tunicates, fish, fish eggs, Mollusca, Amphipoda, Cnidaria, Euphausiacea, Polychaete, Echinoidea, and algae (

Table 4. Prey items of T. bernacchii. are determined by morphological analyses. “*” indicates occurrence, but the corresponding quantity or mass data could not be obtained. F%, N%, and W%, respectively, represent the percentage of frequency of occurrence, quantity, and weight.

Prey Types	F%	N%	W%
Polychaeta	21.74	0.44	8.77
Cnidaria	34.78	2.19	5.36
Amphipoda	13.04	1.02	*
Mollusca
Bivalvia	17.4	*	*
Gastropoda	13.1	0.44	6.53
Echinoidea	30.43	0.44	*
Fish	8.7	0.29	2.49
Tunicata
Ascidiacea	8.7	1.02	17.6
Thaliacea	4.35	0.15	*
Euphausiacea	13.04	2.63	*
Algae	8.7	0.58	9.22
Fish Egg	47.83	90.79	50.02

). Due to varying degrees of digestion among different individuals, it was not possible to obtain the weight of the prey organisms for some of the samples. According to the experimental results, regarding the number of prey items found in the stomach contents, fish eggs were the most common, accounting for 90.79%. This was followed by Euphausiacea and Cnidaria. In terms of mass proportion, fish eggs also had the highest percentage, reaching 50.02%, followed by Tunicates at 17.6%, and then algae and Polychaeta, which accounted for 9.22% and 8.77%, respectively. If considering the frequency of occurrence, fish eggs were most frequently found, recorded in 11 individuals. Next were Cnidaria and Echinoidea, with proportions of 34.78% and 30.43%, respectively. Polychaeta also accounted for 21.74%.

3.2.2. Molecular Analysis of Stomach Contents

Based on the analysis of the DNA metabarcoding results, the stomach contents of Trematomus bernacchii showed a total of 13 phyla, 23 classes, 43 orders, 50 families, 50 genera, and 32 species. Assuming that all these DNAs were the prey of T. bernacchii, by percentage of contigs, Polychaeta accounted for 50.859%, Echinodermata 11.376%, Cnidaria 8.915%%, Crustacea 9.332%, Mollusca 7.670%, and Fish 4.112% (

Table 5. Prey species and relative abundance of T. bernacchii based on metabarcoding results. “Unclassify” refers to OTUs within this percentage that could not be assigned to the order level or any finer taxonomic resolution.

Prey Taxon		Contigs	Prey Taxon		Contigs
Polychaeta		50.860%	Mollusca		7.670%
Phyllodocida	Aglaophamus cf.	5.009%	Bivalvia	Adamussium colbecki	3.983%
	Harmothoe crosetensis	2.636%		elliptica sp1.	0.135%
	Barrukia cristata	1.994%		elliptica sp2.	0.097%
	Harmothoe sp.	1.682%	Gastropoda	Neobuccinum eatoni	2.664%
	Aglaophamus trissophyllus	0.593%		Margarella antarctica	0.373%
	Others	1.202%		Limacina sp.	0.235%
Terebellidae	Amphitrite kerguelensis	15.614%		Marseniopsis sp.	0.056%
	Lysilla sp.	11.468%		Others	0.127%
	Thelepus antarcticus	3.784%	Echinodermata		11.376%
	Others	0.716%	Echinoidea	Abatus agassizii	4.737%
Spionidae	Laonice weddellia	1.596%		Sterechinus neumayeri	0.363%
	Scolelepis eltaninae	0.338%	Asteroidea	Macroptychaster accrescens	4.676%
Chaetopteridae	Unclassify	0.920%		Odontaster sp.	1.462%
Sabellida	Perkinsiana littoralis	0.131%	Asteroidea	Glabraster antarctica	0.119%
Unclassify		3.176%		Others	0.009%
Cnidaria		8.915%	Holothuroidea	Psolicrux sp.	0.004%
Hydrozoa	Corymorpha sp.	4.875%	Unclassify		0.006%
	Obelia sp.	2.441%	Fish		4.112%
	Erenna sp.	0.005%	Nototheniidae	Trematomus sp1.	3.004%
	Bougainvillia sp.	0.004%		Pleuragramma antarctica	0.048%
	Others	1.104%		Trematomus sp2.	0.033%
Anthozoa	Stomphia sp.	0.017%		Trematomus newnesi	0.018%
Scyphozoa	Phacellophora camtschatica	0.331%		Trematomus sp3.	0.004%
Unclassify		0.130%	Channichthyidae	Pseudochaenichthys georgianus	0.108%
Crustacea		9.332%		Chionodraco sp.	0.064%
Amphipoda	Orchomenella pinguides	4.482%	Stromateida	Pampus sp.	0.025%
	Iphimedia sp.	0.715%	Bathydraconidae	Gymnodraco acuticeps	0.228%
	Paramoera walkeri	0.111%	Centrarchidae	Micropterus salmoides	0.412%
	Caprella penantis	0.010%	Unclassify		0.166%
	Others	0.011%	Others		7.734%
Decapoda	Chorismus antarcticus	3.040%	Tentaculata	Unclassify	3.379%
Copepoda	Sinocalanus sp.	0.739%	Pycnogonida	Ammothea clausi	3.079%
	Paracalanus sp.	0.041%		Others	0.074%
	Sinocalanus tenellus	0.010%	Algea	Unclassify	0.175%
	Others	0.021%	Nemertea	Unclassify	0.980%
Euphausiacea	Euphausia crystallorophias	0.012%	Chaetognatha	Sagitta sp.	0.043%
Unclassify		0.140%	Sponge	Unclassify	0.00%

). The DNA T. bernacchii metabarcoding results also indicate that the genes of T. bernacchii itself were present in the stomach, but these were excluded from the analysis, along with species that could not possibly exist in the Ross Sea region.

To determine the taxonomic status of fish egg 1 and fish egg 2, we retrieved the 12S rRNA sequences of species related to fish eggs from the NCBI database and performed a phylogenetic analysis (Figure 2). As shown, Fish Egg 1 and Fish Egg 2 form a highly supported monophyletic group with T. bernacchii (MZ779012.1), indicating a close phylogenetic relationship between them and T. bernacchii.

3.3. Prey Composition Changes with Total Length

To assess changes in prey composition across different body size groups, we focused on the DNA metabarcoding analysis results due to the high degree of digestion observed in the morphological identification process. Prey composition was assessed by analyzing the abundance of OTUs within each size group. The proportion of different prey varied with the increase in body length. Overall, Polychaeta consistently occupied a significant proportion across all body length conditions. When the total length (TL) exceeded 235 mm, T. bernacchii primarily fed on Polychaeta, Echinodermata, and Mollusca. Conversely, when the TL was less than 220 mm, the diet tended to be more inclined towards Cnidaria, Polychaeta, and Crustacea. For Echinodermata, their proportion increased with the increase in TL, while Cnidaria and Crustacea exhibited an opposite trend (Figure 3).

3.4. Prey Composition Changes by Gender

To assess potential differences in prey composition between males and females, we used DNA metabarcoding analysis, with 8 male and 14 female individuals as samples. Comparisons were made by analyzing the abundance of OTUs within each gender. By calculating the proportions of different types of prey in the stomach contents of individuals of different genders, the p-value of the one-way ANOVA is 0.593, which is greater than 0.05. This indicates that there was no significant difference in the proportion of various species groups in the stomach contents between female and male individuals (Figure 4).

4. Discussion

In this study, otoliths were employed to estimate the growth patterns of T. bernacchii. This technique is prevalent in the biological assessment of these species [47], as otoliths can continue to grow even when somatic growth ceases entirely during periods of food scarcity and starvation [48]. Specifically, for T. bernacchii, researchers have utilized sectioned otoliths to ascertain the age and growth rates of these fish [4]. The current investigation determined the maximum age of T. bernacchii to be 19 years, which is closely aligned with the highest recorded age of 21 years in prior studies [49]. These findings corroborate the typical characteristics of Antarctic fish, which exhibit extended growth cycles, slow growth rates, and a propensity for longevity. Such traits have also been validated through tagging experiments [50] and controlled aquarium rearing studies [51].

The results of the stomach content analysis indicate that the primary prey consumed by T. bernacchii was fish eggs. Using DNA barcoding technology, we confirmed that a portion of these eggs were from the same species. The spawning period of T. bernacchii in the Ross Sea ranged from August to January, with a peak between mid-December and mid-January. Therefore, in theory, the possibility of egg cannibalism exists. [52]. Self-cannibalism is prevalent among fish species, and numerous instances of egg cannibalism have been documented in various studies [53,54]. Notably, egg cannibalism has been observed even in species that exhibit parental care [55]. However, the occurrence of egg cannibalism in Antarctic fish is quite rare. Previous studies on T. bernacchii did not report egg-consumption behaviors [53]. However, by the year 2000, instances of egg consumption began to be documented, although eggs remained a minor component of the diet [56]. Concurrently, other Antarctic fish species were also observed to exhibit egg consumption [57]. Using DNA barcoding techniques, egg consumption in T. bernacchii was identified as a form of egg cannibalism. Cannibalism in fish has been increasingly reported in the literature, with ecological implications including its role as a density-dependent regulator of fish populations [58], as well as its potential to enhance the lifetime reproductive success of individuals [59]. Possible drivers of egg cannibalism include food scarcity, post-spawning energy supplementation, or a strategy to improve reproductive success [60,61]. In the present study, egg cannibalism was observed in a female specimen whose gonadal development was at stage six, suggesting that this behavior may serve as a form of energy supplementation following ovulation. From an environmental perspective, in most of Antarctica, sea ice increased between the 1970s and 2014. This increase in sea ice likely led to a reduced flux of primary production to the benthos, consequently affecting the Benthic ecosystem and the organisms inhabiting it [62]. However, the Antarctic environment is highly diverse, and even within the same region, such as Terra Nova Bay in the Ross Sea, Southern Ocean, Benthic ecosystems can vary significantly across different locations. Therefore, more detailed environmental data are needed to investigate how environmental changes may lead to shifts in feeding behavior [63]. Thus, further evidence is needed to investigate the reasons for egg cannibalism occurrence and its consequences. Additionally, the results of the stomach content analysis reveal that Cnidaria, Echinoidea, and Polychaeta also appear frequently. These taxa have been previously reported as typical species of Benthic ecosystems [64]. This is also consistent with previous studies on the feeding habits of T. bernacchii, where these prey categories were found to occur [65]. This result is further consistent with previous studies on T. bernacchii, which suggest that larger individuals tend to consume more Polychaeta [1].

Molecular analysis reveals that Polychaeta constitutes the highest proportion of prey types at 50.859%, followed by Echinodermata (11.376%) and Cnidaria (8.915%). Additionally, there were systematic changes in the diet of T. bernacchii with increasing body length. Primarily, Polychaeta remained the most prevalent across all sizes without significant variation. In contrast, the proportions of Crustacea and Cnidaria gradually decreased with increasing body length. Conversely, the proportion of Echinodermata was positively correlated with body length, while Mollusca were more abundant in larger specimens and less so in smaller ones. This pattern of dietary shift with size is also commonly observed in other Antarctic fishes [66]. Additionally, there were differences in dietary compositions between genders, even though they belong to the same species. The higher occurrence of Echinodermata in female samples and the increased presence of Cnidaria in males could potentially be attributed to variations in body length between the sexes.

Comparing the two methodologies, we found that the morphological approach has the advantages of being straightforward and intuitive. However, these advantages are heavily dependent on the availability of substantial labor and high expertise in species identification. Furthermore, this method has limitations in providing insights into long-term dietary patterns and is constrained by the advanced state of prey digestion often encountered. Conversely, meta-barcoding overcomes these challenges by offering high precision, enabling accurate identification at the genus or species level. Nevertheless, a limitation in our study arose as numerous fish eggs detected in stomach content analysis could not be verified through DNA metabarcoding. Moreover, this method could not distinguish whether the gene fragment was derived from fish prey or fish eggs. Additionally, the Tunicates identified through morphological analysis were not reflected with the same level of significance in the DNA metabarcoding results. This limitation underscores the current shortcomings in available techniques, indicating the need for the further refinement of universal primers. Despite these constraints, our analysis of stomach contents successfully identified algae as relatively intact specimens, which allowed us to rule out indirect ingestion. However, it remains unclear whether the algae were actively consumed or ingested accidentally during the consumption of other prey, requiring further research to clarify this. Overall, the use of both methodologies provided a comprehensive understanding of the feeding habits of T. bernacchii, contributing to a deeper knowledge of its trophic ecology and the structure of the Benthic food web in the Ross Sea.

In this study, we investigated the feeding habits of T. bernacchii; the categorization of this species’ feeding behavior remains incomplete, particularly regarding egg cannibalism, which has yet to be fully elucidated. Furthermore, the sample size may limit the applicability of the findings, and incorporating seasonal, geographical, and inter- and intra-population variations would provide more comprehensive results and better capture the unique biogeographical characteristics of different Antarctic marine environments. The excessive digestion of prey in the stomach hindered the use of stomach content analysis to reflect age-specific prey diversity. Nevertheless, despite these limitations, the study provides contributions to our understanding of the age and feeding habits of T. bernacchii. More work is needed to further expand our understanding of the feeding habits of T. bernacchii.

5. Conclusions

This study provides significant insights into the growth patterns and feeding ecology of Trematomus bernacchii, contributing to a broader understanding of Antarctic fish biology. The determination of a maximum age of 19 years, consistent with prior findings, reaffirms the characteristic slow growth and longevity of Antarctic fish. The stomach content analysis highlights the unique dietary preferences of T. bernacchii and reveals the systematic dietary shifts associated with body size and gender, reflecting complex trophic interactions within the Ross Sea Benthic ecosystem. The comparison of morphological and molecular methodologies underscores the complementary strengths of these approaches. Moreover, this study identifies the rare occurrence of egg cannibalism, predominantly involving conspecific eggs. Overall, these findings enhance our understanding of the ecological role of T. bernacchii and provide a foundation for the conservation and management of Antarctic marine biodiversity.