The global fish production industry plays a crucial role in national economies, supporting an estimated 59.5 million jobs in the primary sector of capture fisheries and aquaculture [1]. Dealing with the most valuable traded food commodity worldwide, seafood has also become a fundamental income product for developing countries with net exports valued more than sugar, tobacco, meat and rice combined [1,2]. Staggering numbers highlight a constant worldwide increase both in the sector of natural seafood capture and in aquaculture production. According to available data, global fish production has reached almost 300 million tonnes [3], also considering world aquaculture, which accounts for about one third of total fish production [1]. In 2018, aquaculture fish production was dominated by finfish (54.3 million tonnes—47 million tonnes from inland aquaculture and 7.3 million tonnes from marine and coastal aquaculture); molluscs, mainly bivalves (17.7 million tonnes) and crustaceans (9.4 million tonnes) [1].

However, if there is production, there is also consumption. In fact, according to the FAO [1], global food fish exploitation increased at an average annual rate of 3.1% from 1961 to 2017, a rate almost twice than the annual world population growth (1.6%) for the same period, and higher than all the other animal protein foods (meat, dairy, milk, etc.), which increased by 2.1% per year. At the individual level, global fish consumption rose by 122% from 1990 to 2018 [1]. The annual per capita seafood consumption of fisheries and aquaculture products approximately doubled in 2018 compared to the level in the 1960s. In particular, per capita fish consumption grew significantly from 9.0 kg in 1961 to 20.5 kg in 2018, by about 1.5% per year [1]. Europeans consume, on average, 24.4 kg per person of fishery products annually, 4 kg more than the world average [4]. Despite persistent differences in levels of fish consumption between regions and states, significant trends and trajectories were observed [5]. All the above cited data are graphically illustrated in Figure 1.

The rise of international fish trade leads to the need to improve the traceability of fishery products. In relation to this, innovation must be introduced at technological level to support the consistent production increase, with a special focus on the monitoring chain assessing product handling, processing and distribution via global networks [6]. A wide number of fish species are nowadays commercialized for human consumption on a world scale, most of which derive from aquaculture production and fishery activities [7]. On the other hand, cultural improvements and attention from the media has led consumers to demand more comprehensive and precise information on fish labelling. Therefore, issues concerning food quality and safety have recently become crucial points, also considering the still-frequent habit of fish species substitutions under certain conditions.

In fact, species substitution happens to expand profits; higher value species are replaced with other less precious, often cheaper, less well-known or even illegal and protected species [8,9]. Fish traceability is nowadays fundamental to avoid substitutions that may carry hidden risks for consumers; basic consequences may be health problems that occur primarily through the consumption of cryptic species coming from contaminated areas without any sanitary checks or able to trigger allergy problems [10,11]. It must be remarked that species substitution might even occur accidentally when taxa are difficult to recognize at a morphological level, and consequently the systematics of closely related species are trivial [4].

The European Union (EU), within the renewal plan of the Common Fisheries Policy and the Common Market Organization, has introduced new requirements for the labelling of fisheries and aquaculture products through the Cape IV of Regulation EU No. 1379/2013. Although this regulation requests appropriate species traceability and labelling (scientific binomial nomenclature based on genus and species collectively with the common name), the identification of processed species is frequently difficult to perform. Morphology-based identification methods may lead to incorrect species identification. Nowadays, more innovative methods and technologies are used to assess taxa determination and authenticity. Molecular diagnostic techniques have been developed to identify food fraud using different approaches, from the use of single proteins through enzyme-linked immunosorbent assay (ELISA) [12] or species-specific DNA sequences (DNA Barcode) [13] to near infrared spectroscopy [14] or modern genomic approaches [15].

In relation to the above cited issues, the aim of this review is to highlight the main barcoding approaches to identify the most reliable ones, which are able to allow affordable taxonomic identification of cryptic seafood species and limit commercial fraud that might threaten the consumer’s health or the survival of endangered species.

The review has been prepared searching mainly in “ResearchGate”, “Google Scholar”, “PubMed”, “Scopus” and “Web of Science”. Considering that a high-quality search cannot rely on a single database, inclusion of different datasets is helpful to obtain more reliable literature. High-impact journals were preferred to avoid scarce quality of data and, above all, to have a wider audience.

In recent years, molecular barcoding (DNA barcoding) has been suggested as the gold standard in forensic taxonomy [16] and can be considered a further development of the previously applied technique proposed by Bartlett and Davidson [17]. About 30 years ago, these scientists presented an innovative methodology directed at cryptic food identification based on the use of short nucleotide regions for species authentication called Forensically Informative Nucleotide Sequence (FINS). This technique consists of a specific segment of DNA amplified using PCR and combines DNA sequencing and phylogenetic analysis to identify the most closely related taxon. Despite that, the modern concept of DNA barcoding was not established yet; the actual idea of DNA barcoding was developed in 2003 by Hebert et al. [18]. DNA barcoding is the analysis of variability in a specific genomic region, which is therefore designated the “DNA barcode”, to be compared with specific databases of previously analysed sequences that become a priori reference DNA fragments determined for the species of interest [3]. The predominant precept driving DNA barcoding is the amplification of homologous genes by means of PCR and subsequent DNA sequencing. Sequences of interest are used as a “barcode” to determine the identity and authenticity of food products, for example, for DNA identification of various plant and animal species, to discriminate the presence of different taxa in processed food and to assess the presence of raw materials in food industry processes [19]. DNA barcodes consist of a standardized short sequence of DNA (in the range 400–800 bp) that, in theory, should be easily generated and characterized for all species [20]. The central notion of DNA barcoding asserts that a short sequence of DNA should display low variability within species and greater differentiation between species [20]. Kress and Erickson [21] proposed three criteria that have to be satisfied to consider a gene region as a DNA barcode: (i) contain significant species-level genetic variability and divergence, (ii) possess conserved flanking sites for developing universal PCR primers for wide taxonomic application and (iii) have a short sequence length to facilitate current capabilities of DNA extraction and amplification.

Nowadays, although both mitochondrial (mtDNA) and nuclear (nDNA) genes are involved in variegated approaches, the most reliable barcodes for the discrimination of different animal species are obtained using the mitochondrial gene coding for cytochrome c oxidase 1 (COI) and cytochrome b (cytb) [18,22,23,24,25]. In particular, the most-used DNA barcode for seafood identification is a ~650 bp fragment of the mitochondrial gene COI. Many studies have shown the applicability of COI barcoding for accurate identification of a wide range of fish species and mislabelling detection [24,26,27,28,29,30].

It must be remarked that the use of mitochondrial markers for species’ correct taxonomy displays several advantages with respect to nDNA (Table 1). In particular, mtDNA has a matrilinear inheritance and is not subjected to recombination. For this reason, nucleotide variation within the same taxon is at a minimum level while, in nDNA, great differentiation emerges among different species [3]. In fact, in spite of the high mutation rate of some mitochondrial regions, COI and cytb are conserved genes within each species.

Another advantage can be referred to the high number of copies of mtDNA contained in the same specimen that allows a more reliable amplification in case of degraded samples such as those derived from processed seafood. It is noteworthy that DNA redundancy is generated by several contemporary copies of mtDNA within the same tissue; attention must be dedicated to the correct evaluation of data, and alternative analyses using nDNA might also be considered.