The importance of distinguishing pufferfish species (Lagocephalus spp.) in the Mediterranean Sea for ensuring public health: Evaluation of the genetic databases reliability in supporting species identification

Abstract

Taxonomic identification of marine organisms is sometimes hindered by morphological similarities and utilization of wrong criteria. Therefore, the morphological approach often requires the support of molecular tools which usually rely on a comparison of DNA sequences available in free publicly-accessible databases. However, the process can be affected by wrongly deposited sequences which lead to specimens’ misidentification. This is the case of two toxic pufferfish species (Lagocephalus spadiceus and L. guentheri), both reported as Lessepsian invasive species, whose actual presence in the Mediterranean is debated within the scientific community. In this study, the reliability of the genetic databases GenBank and BOLD in supporting the discrimination of L. spadiceus and L. guentheri was assessed as it has been already debated in literature. Twenty Mediterranean specimens of L. guentheri were collected and morphologically identified. COI and cytb reference sequences were then produced and included in two separate analyses (one for each gene) together with corresponding online sequences of L. spadiceus and L. guentheri from all the available localities. A high percentage of sequences with non-valid taxonomic identification was observed, involving 32.5% of the COI and 43.7% of the cytb sequences from GenBank and 30% of the COI sequences from BOLD. The majority of sequences deposited under L. spadiceus, mostly of Mediterranean origin, were genetically confirmed to be misidentified L. guentheri. Outcomes highlighted two main shortcomings: i) a low taxonomic accuracy of official databases due to the presence of sequences attributed to wrong species ; (ii) a significant underestimation of L. guentheri presence in the Mediterranean Sea. This study, therefore, underlines the necessity to improve the databases accuracy in term of deposited sequences reliability. In this specific case, accuracy is even more important, considering the involved toxic species and the potential concern for public health associated with their accidental entering in the seafood chain.

Introduction

According to the last update of the World Register of Marine Species database - WoRMS (http://www.marinespecies.org/index.php), more than 240,000 accepted species names are known to be distributed in the marine environment. Acquainting with biodiversity and identifying species is fundamental to sustain the marine ecosystem, guard against worldwide threats such as climate change, pollution, overfishing, illegal fishing, habitat destruction, and invasive species (Duffy et al., 2013; Coll et al., 2014; Bunholi et al., 2018). It even acquires fundamental importance in the prevention of phenomena of counterfeiting and adulteration that entail economic loss for consumers and potentially have a damaging effect on public health if toxic species accidentally enter in the seafood chain. Species identification usually requires the expertise of taxonomists and has been conventionally based on morphological observations and measurements. However, in some cases these practices are inadequate for their purpose (Wheeler et al., 2004). For instance, the morphological approach may not be applicable to species identification in early life history stages (eggs and larvae), as well as to the detection of cryptic species (Pampoulie and Daníelsdóttir, 2008; Trivedi et al., 2016).

With this respect, DNA-based methods are nowadays increasingly used to reveal new non-indigenous species, as well as serving to accurately identify and detect species in seafood products. Sequencing techniques are currently the most applied and, as a rule, rely on comparisons of DNA sequences from publicly available reference libraries. GenBank (https://www.ncbi.nlm.nih.gov/genbank/) is a comprehensive database that contains nuclear and mitochondrial nucleotide sequences for more than 400,000 named organisms, obtained primarily through submissions from individual laboratories and batch submissions from large-scale sequencing projects (Benson et al., 2018). Moreover, the Barcode of Life Data system - BOLD (http://www.boldsystems.org/) has gained worldwide popularity with the development and the success of DNA Barcoding, based on the use of a partial sequence of the mitochondrial cytochrome c oxidase subunit 1 (COI) gene as a target region for species identification and discrimination (Hebert et al., 2003), and includes nowadays sequences from more than 270,000 organisms.

However, pitfalls in both online-depositories are related to the fact that many of the available sequences lack validation in the form of voucher information (Zanzi and Martinsohn, 2017). Therefore, DNA sequences obtained from specimens that were not properly vouchered may belong to misidentified species, leading to a worrisome decrease in database reliability and to a consequent unavoidable presence of gaps and imprecisions in the knowledge of the marine environmental status. The spreading of the invasive alien pufferfish species in the Mediterranean Sea initiated an emerging concern for public health, considering their poisonous viscera and flesh (Rambla-Alegre et al., 2017). In fact, they naturally harbour a heat-stable neurotoxin, called tetrodotoxin (TTX), which is potentially lethal if ingested in sufficient quantities (Bane et al., 2014). Nevertheless, pufferfish are known to possess varied levels of toxin, thus it is important to confirm their exact identification for assessing and managing the risk related to their accidental fishing and subsequent entering in the seafood chain. With this regard, previous studies have especially highlighted problems in discriminating two pufferfish species from the genus Lagocephalus: L. spadiceus and L. guentheri (Miranda Ribeioro, 1915), both morphologically and genetically similar (Vella et al., 2017).

The available scientific literature has reported them as invasive ‘Lessepsian species’ that is originated from the Red Sea (El-Haweet et al., 2016; Farrag et al., 2016; Turan et al., 2017). Considering the morphological resemblance between the two species and their recent revision (Matsuura et al., 2011; Psomadakis, 2015), it has been hypothesized that many of the old Mediterranean records of L. spadiceus could have been erroneously identified. The debate has gone so far that the real occurrence of L. spadiceus is reputed to be questionable (Vella et al., 2017). Moreover, although both the species are reported as “harmful” on the official finfishes’ database (www.fishbase.org), L. spadiceus is described in literature as occasionally poisonous (Chulanetra et al., 2011) and therefore represents an actual health risk if accidentally by-caught during commercial fishing. In this light, solving the issues hindering the proper discrimination between the aforementioned species might be essential for both clarifying the actual status of Mediterranean Sea in term of toxic species presence and assessing the consumers’ risk exposure.

In this study, the accuracy of GenBank and BOLD databases regarding the deposited sequences of L. spadiceus and L. guentheri was in-depth evaluated. The taxonomic classification of the specimens from which the public sequences belong was revised after performing a phylogenetic analysis using reference sequences of L. guentheri produced in this study and problems related to database accuracy in identification of toxic species were discussed. Moreover, the Mediterranean records of the two species were re-evaluated in the light of the molecular results.