MICROSATELLITE DNA-BASED GENETIC TRACEABILITY OF TWO POPULATIONS OF SPLENDID ALFONSINO , BERYX SPLENDENS ( ACTINOPTERYGII : BERYCIFORMES : BERYCIDAE ) — PROJECT CELFISH — PART 2

Background. The study is a contribution to Project CELFISH which involves genetic identifi cation of populations of fi sh species presenting a particular economic importance or having a potential to be used in the so-called commercial substitutions. The EU fi sh trade has been showing a distinct trend of more and more fi sh species previously unknown to consumers being placed on the market. Molecular assays have become the only way with which to verify the reliability of exporters. This paper is aimed at pinpointing genetic markers with which to label and differentiate between two populations of splendid alfonsino, Beryx splendens Lowe, 1834, a species highly attractive to consumers in Asia and Oceania due to the meat taste and low fat content. Material and methods. DNA was isolated from fragments of fi ns collected at local markets in Japan (MJ) (n = 10) and New Zealand (MNZ) (n = 18). The rhodopsin gene (RH1) fragment and 16 microsatellite DNA fragments (SSR) were analysed in all the individuals. The sequences obtained were processed using the BioEdit and BLAST software, whereas SSR data were processed with the GeAlEX analysis package. Results. The BioEdit software-aided comparison of MJ and MNZ nucleotide sequences of the rhodopsin gene fragments were identical and showed 100% agreement with the alfonsino sequence deposited under access number DQ197832. The preliminary analysis of SSR markers showed all the loci analysed in both populations to be polymorphic, and when randomly selected specimens were assigned to the original populations. The affi nity test correctly identifi ed the provenance of all those specimens. Conclusion. The results obtained constitute a tool for molecular differentiation between alfonsino populations collected in the FAO 81 (New Zealand) and FAO 71 (Japan) areas for the purpose of catch quota control and for checking the agreement between the label declaration and the actual product.


INTRODUCTION
As consumers in the developed countries require accurate information on the goods available on the market, detection of food counterfeiting has become an increasingly important issue.The issue is particularly relevant to fi sh and shellfi sh as it is diffi cult to authenticate products of fi sheries and aquaculture.In many cases, there are no methods with which to effi ciently verify the identity of such products.Consumers are increasingly frequently prone to select goods of 'certifi ed origin' and do not wish to purchase fi sh from illegal catches.Environmental issues affect those choices as well.Today's consumers require information whether the open sea products available on the market have been harvested without harming populations, causing overfi shing through selection of inappropriate sizes, or fi shing the species regarded by IUCN as threatened.According to Johnson (2014), fi sh and shellfi sh counterfeiting primarily involves substitution of valuable species by less nutritionally benefi cial ones which in turn may be a source of gastric infections or unexpected allergies.The latter are particularly dangerous for children and seniors as well as for individuals on diets that exclude consumption of certain types of animal fat and/or protein.In such cases, food counterfeiting poses a health danger and should not be tolerated.
During three years (2011)(2012)(2013), the West Pomeranian University of Technology in Szczecin (Poland) has been involved in a research carried out within the framework of the Project CELFISH ("Development of a genetic-based system for identifi cation of food products form fi sheries and aquaculture introduced to the European Union customs area") (See Part 1 of this project contribution at DOI: 10.3750/AIP2014.44.2.08).The project participants have collected about 12 000 samples of fi sh from several countries worldwide which contribute most to the European Union seafood imports.The molecular assays performed are meant to provide the customs and food inspection services with a DNA-based tool with which to identify not only a fi sh species, but a population and the capture site.The project will deliver a genetic data base, the development of which is in progress.The data base will cover about 300 fi sh species from waters off all the continents.Genetic traceability of fi sh products as well as population identifi cation using molecular methods provide useful information about the catch region and is important for developing conservation plans for overfi shed populations or catches from IUU (Illegal, Unreported, and Unregulated) fi sheries.
The splendid alfonsino, Beryx splendens Lowe, 1834 (thereafter alfonsino), is a widely distributed species occurring in North Atlantic (from the Gulf of Maine to the Gulf of Mexico), East Atlantic (from off south-western Europe and the Canary Islands to South Africa), and the Indo-Pacifi c region (from off East Africa, including Saya de Malha Bank, to Japan, Hawaii, Australia, and New Zealand).The limited number of records in the western Pacifi c is doubtless a result of limited fi shing effort at depths exceeding 200 m.The splendid alfonsino (later in the text referred to as alfonsino) is absent only from the Mediterranean Sea and the north-eastern Pacifi c (Froese and Pauly 2016).Adult fi sh inhabit mainly the outer shelf (180 m) and the slope to the depth of at least 1300 m, occurring mostly between 200 and 600 m; the species is often found above seamounts (Paxton 1999) and underwater ridges (Dubochkin and Kotlyar 1989).The alfonsino is oviparous and spawns in batches (Lehodey et al. 1997).Beryx splendens feed mainly on fi sh, crustaceans, and cephalopods (Dubochkin and Kotlyar 1989).According to taxonomic studies carried out by Johnson and Patterson (1993), the families Berycidae and Holocentridae and the suborder Trachichthyoidei form a monophyletic group.Extensive information on the beryciforms has been supplied by Kotlâr (1996) who reported the order to consist of 7 families with 29 genera and 144 species.The alfonsino life span is assessed to average several years, the oldest reported specimen being 23 years old (Adachi et al. 2000).
The nature of alfonsino's migrations (unrestricted or spatially constrained) is not known, nor has the gene fl ow between stocks from various fi shing grounds been followed.According to Lévy-Hartmann et al. (2011), the alfonsino shows an extremely high intraspecifi c diversity, which suggests the presence of genetically separate stocks or even populations.The phylogenetic diversity index (PD50) was estimated at 0.6260 (Faith et al. 2004), which, on the scale of 0.5 (low) to 2.0 (high) indicates a low inter-population diversity, a high diversity being indicated within lower taxonomic groups (stocks or subpopulations).
The annual landings of alfonsino in New Zealand and Korea amount to 2900 and 560 t, respectively; the combined landings in Spain, Chile, and Portugal adding up to about 200 t (Anonymous 2013).The alfonsino fi sheries in New Zealand, Australia, and Chile are controlled by catch quota (Bensch et al. 2009).On account of its high meat quality, the species is of a paramount commercial importance and is exported mainly as frozen fi sh.According to the IUCN threat status established based on comprehensive ecological and biological research, the species falls into the category of 'not evaluated', which also shows that the biology, autecology, and genetics of the species merits study.On the other hand, the intrinsic extinction vulnerability assessment of Cheung et al. (2005) classifi es the species as vulnerable.Other sources stress the necessity of determining the intraspecifi c variability of the alfonsino (Bensch et al. 2009).So far, information on population genetics of this species is scarce, and these result published by Lévy-Hartmann et al. ( 2011) must be updated.The only data on the cytogenetic index have been provided by Japanese workers (Ojima and Kikuno 1986) and concern the number of chromosomes in diploid cells (2n = 48) and gametes (n = 24).In 2006, Chilean scientists who analysed the alfonsino's parasitic fauna found that there most probably exist two non-mixing stocks.However, the amount of available data was regarded as too low for making inferences on the population status (Niklitschek et al. 2007).

RESULTS
The BioEdit software-aided comparison of MJ and MNZ nucleotide sequences of the rhodopsin gene fragments showed no differences.Subsequently, the RH1 sequences were compared, using the BLAST software, with records deposited in the GenBank.The comparison showed the sequences to agree in 100% with the alfonsino sequence deposited under access number DQ197832.The preliminary analysis of SSR markers showed all the loci analysed in both populations to be polymorphic, so they were deemed suitable for further analyses.The total number of alleles in the MJ and MNZ populations was 175 and 244, respectively.The number of alleles per locus ranged from 2 to 18 (mean: 10.94) and from 2 to 24 (mean: 15.25) in the MJ and MNZ material, respectively.The allele patterns in both populations are shown in Fig. 1.The observed heterozygosity (Ho), i.e., the ratio between heterozygotic genotypes and the total number of genotypes per locus, was found to range between 0.2 and 1.0 (mean: 0.844) and between 0.056 and 1.0 (mean: 0.837) for the MJ and MNZ material, respectively.The genetic diversity (He) was found to range within 0.180-0.935(mean: 0.829) and within 0.054-0.943(mean: 0.838) for the MJ and MNZ material, respectively (Table 1).No signifi cant deviation from the Hardy-Weinberg equilibrium was found in all the loci analysed in the MJ population and in 14 loci analysed in the MNZ one.It was only the MNZ's Orla 2-91 and Orla 3-151 loci that showed a signifi cant (P < 0.05 and P < 0.01, respectively) deviation from the Hardy-Weinberg equilibrium (Table 2).When randomly selected specimens were assigned to the original populations, the affi nity test correctly identifi ed the provenance of all those specimens.The test was carried out at the 0.05 signifi cance level (Fig. 2).

DISCUSSION
Genetic identifi cation of two clades corresponding to populations of the alfonsino caught off Japan and off New Zealand indicates their biological intraspecifi c separation.Attempts to defi ne the geographic distribution-based structure of the species had been earlier made by Hoarau and Borsa (2000).Having analysed the cytochrome b gene sequence they demonstrated that the species Beryx splendens sensu lato consists of two sub-species, the within-sub-species variation being relatively low.Such genetic homogeneity of lower taxa may be indicative of the absence of gene fl ow, which in turn is conducive to separation of systematic structures such as stocks, sibling species or ecological species.
The alfonsino is particularly vulnerable to population size reduction by, e.g., overfi shing.Catch records of other fi sh, such as bluenose warehou, Hyperoglyphe antarctica (Carmichael, 1819); silver gemfi sh, Rexea solandri (Cuvier, 1832); blue grenadier, Macruronus novaezelandiae (Hector, 1871); and thorntooth grenadier, Lepidorhynchus denticulatus Richardson, 1846 contain records of alfonsino individuals as well (Bensch et al. 2009).However, no gear or fi shing technique contribution to knocking the alfonsino population off balance has been determined with any precision.As indicated in this paper, research on the  species' genetic status is necessary; such research will make it possible to take measures for developing an alfonsino management plan and for setting catch quota.
Owing to the economic importance of the alfonsino, particularly in New Zealand and in Asian countries, and considering the high meat quality of the species, resources represented by individual populations should be estimated and the populations requiring protective measures should be identifi ed.The problem of genetically homogenous populations does not involve the restriction of the gene pool only and disruption of the Hardy-Weinberg equilibrium.It primarily entails catching of too many juvenile individuals.The market price of the alfonsino meat ranges from NZD 6.50 (whole fi sh) to NZD 23 (skinless fi llet) per kilogram.This is the price range groupers, snappers, and turbot that was in 2013 on the pricelist of the Wellington Trawling Company * .In Japan, the whole fi sh price is JPY 1200 (about USD 10) per kilogram.The profi t motive may lead to substituting the species, particularly when sold as half-products, with less expensive fi sh of a lower nutritive value, e.g., the common warehou, Seriolella brama (Günther, 1860).Genetic markers are being developed to detect such substitutions which mislead consumers by offering them cheaper and less valuable product than that declared on the label.Genetic markers offer a possibility, based on the genome analysis, to authenticate both the species and the capture site of valuable and commercially important fi sh species.As shown by this study, it is within reach to check whether the fi sh were caught off Japan or off New Zealand.This in turn makes it possible to identify the provenance of individual batches and, ultimately, will allow revealing how catch quota regulations are observed in those countries where seafood management is not yet perceived as an indispensable component of resource conservation.Subsequently, the approach will make it possible to rebuild the population size and provide a means for controlled exploitation.

ACKNOWLEDGEMENTS
The presently reported study is the second part of a larger research project carried out within 2011-2014 and entitled "Development of a genetic-based system for identifi cation of food products from fi sheries and aquaculture introduced to the European Union customs area".

n
= number f specimens studied, Na = number of different alleles, Ne = number of effective alleles, I = Shannon Information Index, Ho = observed heterozygosity, He = expected heterozygosity, uHe = unbiased expected heterozygosity, F = fi xation index; SE, standard error.

Table 1
Mean genetic diversity of the Japanese (MJ) and New Zealand (MNZ) populations of splendid alfonsino, Beryx splendens, as revealed by the analysis of 16 SSR loci