The biodiversity of the Mediterranean Sea is currently threatened by the introduction of Non-Indigenous Species (NIS). Therefore, monitoring the distribution of NIS is of utmost importance to preserve the ecosystems. A promising approach for the identification of species and the assessment of biodiversity is the use of DNA barcoding, as well as DNA and eDNA metabarcoding. Currently, the main limitation in the use of genomic data for species identification is the incompleteness of the DNA barcode databases. In this research, we assessed the availability of DNA barcodes in the main reference libraries for the most updated inventory of 665 confirmed NIS in the Mediterranean Sea, with a special focus on the cytochrome oxidase I (COI) barcode and primers. The results of this study show that there are no barcodes for 33.18% of the species in question, and that 45.30% of the 382 species with COI barcode, have no primers publicly available. This highlights the importance of directing scientific efforts to fill the barcode gap of specific taxonomic groups in order to help in the effective application of the eDNA technique for investigating the occurrence and the distribution of NIS in the Mediterranean Sea.

Biomonitoring, COI, DNA barcode, environmental DNA – eDNA, gap analysis, Mediterranean Sea, Non-Indigenous Species – NIS, reference libraries

The Mediterranean Sea represents one of the most important biodiversity hotspots in the world (Myers et al. 2000; Coll et al. 2010; Lejeusne et al. 2010; Marrocco et al. 2019), accounting for more than 17,000 reported marine species (Coll et al. 2010). However, the number of Non-Indigenous Species (NIS) and their impact on native species is steadily increasing (Villèle and Verlaque 1995; Streftaris and Zenetos 2006; Marrocco et al. 2018; Bariche et al. 2020). Therefore, the Mediterranean Scientific Community highlights the importance of early warnings and monitoring the presence and distribution of NIS (Katsanevakis et al. 2016; Darling et al. 2017; Tsiamis et al. 2020). Nowadays, this is a key requirement for the conservation and management of ecosystems, as stated by the regulation (EU) No 1143/2014 of the European Parliament and of the Council of 22 October 2014 on the prevention and management of the introduction and spread of invasive alien species (Tiralongo et al. 2019).

Until now, the assessment of Mediterranean species diversity has been carried out through traditional methods based on sole morphological identification. These methods present several disadvantages, such as the difficulty in surveying large geographical areas and spotting and identifying the so-called “hard-to-detect species” (Tiralongo et al. 2020). Besides, they are mainly based on recognizable adult features, such as the shape of gonads or other particular body parts and, often, do not give any identification key for the larval forms or the early developmental stages (Ponti et al. 2009; Di Sabatino et al. 2014; Pinna et al. 2017). They also easily mislead the identification of individuals when their morphology is altered by stressful environmental conditions or by sampling and preservation techniques (Leese et al. 2016; Pawlowski et al. 2018; Tiralongo et al. 2020). In addition, traditional phenotypic-based methods require the expertise of taxonomists, especially when there is the need to identify a species never observed before in a certain area, including NIS (Leese et al. 2016; Pawlowski et al. 2018).

Consistent biological records can provide a better understanding of the distribution of marine species, their expansion range, and the arrival of new NIS in the Mediterranean basin (Mannino et al. 2019; Bariche et al. 2020). A promising approach for the identification of species and biomonitoring of ecosystems is the use of molecular tools such as DNA barcoding, metabarcoding and environmental DNA (Pawlowski et al. 2018; Specchia et al. 2020; Pinna et al. 2021; Tzafesta et al. 2021). DNA barcoding refers to a single species identification with the use of a short DNA fragment, while in the metabarcoding technique the DNA is extracted from a sample containing more than one organism/species, amplified and sequenced by Next Generation Sequencing (NGS) (Ji et al. 2013; Deiner et al. 2017). DNA metabarcoding allows the identification of species at low densities and the detection of taxa that traditional approaches generally fail to distinguish (Pawlowski et al. 2018; Zangaro et al. 2020). Another innovative technique for the identification of species at low concentration is the eDNA (Pawlowski et al. 2018), which is based on the extraction of DNA directly from environmental samples like water or sediment (Rees et al. 2014). This technique can efficiently be applied for assessing the presence and the distribution of NIS that are hard to detect. Also, it may result in a perfect tool for monitoring and preventing new NIS arrivals, for instance, by analysing the genetic content of ballast waters and monitoring it in water exchanges between different regions (Bariche et al. 2020; Tzafesta et al. 2021).

However, even the application of molecular techniques faces some challenges. The level of uncertainty linked to eDNA for marine environments depends generally on the persistence time of the DNA in marine systems (Collins et al. 2018), the high level of connection and movement due to the aquatic medium, and the incompleteness of the reference public libraries (Cagnacci et al. 2012; Weigand et al. 2019; Specchia et al. 2020). Once a DNA fragment has been sequenced, it needs to be blasted into reference libraries to identify the species it belongs to. Moreover, the success of the eDNA technique is also based on the efficiency of the primer sets on large numbers of taxa, a key requirement to correctly amplify the investigated gene in the environmental sample to identify as many species as possible (Elbrecht et al. 2017).

The main DNA barcode reference libraries are GenBank, by the National Centre for Biotechnology Information (NCBI), and BOLD (Barcode of Life Data) Systems (Ratnasingham et al. 2007; Leese et al. 2016; Macher et al. 2017). The information available in the reference libraries includes the species name, the nucleotide sequence of the target genes and, optionally, the PCR primer pairs used for the amplification of the gene of interest in the target organism (Ratnasingham et al. 2007; Macher et al. 2017).

The Consortium for the Barcode of Life (CBOL; www.barcoding.si.edu.com) and the International Nucleotide Sequence Database Collaborations (INSDC) designated the mitochondrial cytochrome oxidase subunit I (COI) as the main barcoding gene based on its widespread presence among different taxonomic groups (Hebert et al. 2004a; Hebert et al. 2004b; Saunders 2005; Ward et al. 2005). Moreover, nucleotide sequence polymorphisms of this approximately 500 bp COI barcode region provide valuable information not only on species identification but also on population genetic diversity and structure (Goetze et al. 2016; Abbas et al. 2018; Choo et al. 2020). Several barcoding studies have also identified alternative genes that can be successfully used for molecular barcoding and may be more suitable to a specific taxonomic group. For example, the ribosomal genes 16s and 18s are generally used for the identification of prokaryotes (Stackebrandt 1994; Acina et al. 2004) and eukaryotes (Hadziavdic et al. 2014; Bradley et al. 2016), respectively; the nuclear ribosomal internal transcribed spacer 1 and 2 (ITS) for fungi (Scoch et al. 2012; Badotti et al. 2017); and two plastid genes, the maturase-coding gene (matK) and the large subunit of ribulose 1,5-bisphosphate carboxylase-coding gene (rbcL) for plants (CBOL 2009), among others.

In light of this, we wanted to evaluate the current status of DNA barcode availability for the NIS already detected in the Mediterranean through morphological surveys. To do so, we retrieved the most recent list of NIS published by Zenetos and Galanidi (2020) and we looked for the availability of COI barcodes and primers in reference libraries. If COI barcodes were not retrieved, we then searched for other barcoding genes. If no records were found in the reference libraries, this was then referred to as DNA barcode gap.

The aim of this research is to evaluate the current limitations in the application of molecular barcoding due to the barcode gap of Mediterranean NIS, and to investigate in depth the occurrence of COI gene barcode and primer pairs. Furthermore, we indicate which taxonomic groups may be underestimated by using molecular tools for the detection and biomonitoring of NIS in the Mediterranean Sea through environmental DNA.

The aim of this study was to quantify the extent of the DNA barcode gap for the NIS established in the Mediterranean Sea, as identified by Zenetos and Galanidi in 2020 and, in doing so, to direct the efforts of the scientific community towards specific taxonomic groups. The data show that 33% of NIS do not have any record in public libraries, making it impossible to detect these species through DNA barcoding techniques. Bryozoa and Parasites are the ones with the largest gap relative to the total number of species in each group. However, Mollusca, which covers almost 25% of the total number of NIS, also needs attention having a barcode gap in 50% of the species, followed by Polychaeta with a gap of 41%. On the other hand, Phytobenthos, Ascidians and especially Fish appear to be the groups that could be mostly identified through molecular techniques, having smaller barcode gaps of 18%, 14% and 3%, respectively.

Our analysis highlighted the importance of analysing barcode gaps in reference libraries for the successful application of molecular tools (including eDNA and DNA metabarcoding) in biomonitoring assessments. Gap-analysis surveys focusing on DNA barcode presence in public repositories for different groups of species are recently gaining greater attention from the scientific community. Gap-analysis has already been applied on marine NIS (Duarte et al. 2021), on macrofauna of a region of the North Sea (Hestetun et al. 2020), on aquatic macroinvertebrates of South-East Italy (Specchia et al. 2020), on marine macroinvertebrates of the Atlantic Iberia (Leite et al. 2020), and on Ascidians and Cnidarians of the European Register of Marine Species (ERMS; Paz and Rinkevich 2021). However, to our knowledge, this is the first study that investigates the DNA barcode gap for NIS occurring in the Mediterranean Sea.

This study also confirms that COI is a useful genetic marker because it is broadly sequenced across different phyla, making it a good candidate gene for identifying species in an environmental sample. Nonetheless, relying only on one DNA fragment may lead to misidentification of pooled samples due to possible sequence similarity; this is why multigene approaches should be preferred in molecular biomonitoring studies (Zou et al. 2012; Chesters et al. 2015; Gangan et al. 2019). Phytobenthos could be the first group to apply this approach, having 30% of species already barcoded with rbcL but not with COI. This is probably explained by the fact that rbcL is a standard barcode for plants (CBOL 2009; Maloukh et al. 2017; Kang et al. 2017; Weigand et al. 2019), further proving the advantage of selecting not only universal barcodes but also relevant taxa-specific genes.

Moreover, the success of the eDNA metabarcoding is based on the availability of efficient primer sets for the amplification of several taxa in a given sample (Elbrecht et al. 2017; Tzafesta et al. 2021). However, only 26% of NIS occurring in the Mediterranean Sea appear to have publicly available COI primers. Especially for animals, 147 species out of 324 COI barcoded species (45.40%) were lacking primer pairs, which highlights the need for further evaluation of primers or the design of new ones. In addition, no universal primer pair was identified, resulting in more laborious molecular identifications where an environmental sample needs to be amplified with several sets of taxa-specific or even species-specific primers to be correctly assessed.

For the above reasons, also the primer pair gap needs to be filled. In order to do so, both increasing the surveys regarding NIS occurring in the Mediterranean Sea and improving the barcoding studies at a global scale is essential, as well as biodiversity assessments (Zangaro et al. 2021). Although the content of the databases doubles approximately every 18 months (https://www.ncbi.nlm.nih.gov/genbank/statistics/), probably many of the NIS established in the Mediterranean Sea come from underdeveloped regions, which cannot financially support molecular surveys. This is easily extrapolated by looking at the “Data Releases” page provided by BOLD Systems (https://www.boldsystems.org/index.php/datarelease). Because of that, we not only encourage collaboration of researchers in this sector, but we also stress the need for training and inclusion of researchers from developing countries, which represent the current and, probably, the future source of new and hard-to-detect NIS.

In conclusion, it is essential to underline that molecular techniques represent a great opportunity to improve the study on the occurrence and distribution of NIS. Hence, a specific gap needs to be filled by the scientific community to make molecular identification totally efficient and independent at a regional, national, and transnational level.

This research was supported by the ex-60% fund from the Italian Ministry of University and Research, by Funding of Basic Research Activities (FFABR) from the Italian Ministry of University and Research (MUR) awarded to M. Pinna and V. Specchia, by the project “Dipartimenti di Eccellenza” awarded to DiSTeBA and by the ImPrEco project funded by Interreg ADRION Programme 2014–2020 awarded to M. Pinna that supported a grant for F. Zangaro.

The authors are grateful to the anonymous reviewers for their useful comments and suggestions.