Establishing and using a genetic database for resolving identification of fish species in the Sea of Galilee, Israel

Abstract

Freshwaters are a very valuable resource in arid areas, such as Mediterranean countries. Freshwater systems are vulnerable ecological habitats, significantly disturbed globally and especially in arid areas. The Sea of Galilee is the largest surface freshwater body in the Middle East. It is an isolated habitat supporting unique fish populations, including endemic species and populations on the edge of their distribution range. Using the Sea of Galilee for water supply, fishing and recreation has been placing pressure on these fish populations. Therefore, efficient monitoring and effective actions can make a difference in the conservation of these unique fish populations. To set a baseline and develop molecular tools to do so, in this study, DNA barcoding was used to establish a database of molecular species identification based on sequences of Cytochrome C Oxidase subunit I gene. DNA barcodes for 22 species were obtained and deposited in Barcode of Life Database. Among these, 12 barcodes for 10 species were new to the database and different from those already there. Barcode sequences were queried against the database and similar barcodes from the same and closely related species were obtained. Disagreements between morphological and molecular species identification were identified for five species, which were further studied by phylogenetic and genetic distances analyses. These analyses suggested the Sea of Galilee contained hybrid fish of some species and other species for which the species definition should be reconsidered. Notably, the cyprinid fish defined as Garra rufa, should be considered as Garra jordanica. Taken together, along with data supporting reconsideration of species definition, this study sets the basis for further using molecular tools for monitoring fish populations, understanding their ecology, and effectively managing their conservation in this unique and important habitat and in the region.

Fig 1. A neighbor-joining unrooted tree of species based on COI haplotypes.

Next to branch edges are haplotype acronyms as listed in Table 2. Numbers next to bifurcations denote percent bootstrap support. For some species, more than one haplotype was identified (denoted A, B etc.). For major families, which are represented by multiple species, a mean K2P between species distance matrix is given to the right. (*)–HmC haplotype derived from specimens suspected as Hypophthalmichthys molitrix x Hypophthalmichthys nobilis hybrids, matches H. nobilis sequences from BOLD. (**)—OaB haplotype derived from a specimen suspected as Oreochromis niloticus x Oreochromis aureus hybrid, matches O. niloticus sequences from BOLD.

Fig 2. Sampling sites and clustering of Garra samples based on COI haplotypes.

Acronyms of populations are: GrA–Garra rufa from haplotype A; GrB–Garra rufa from haplotype B; GrC–Garra rufa from haplotype C; GrD–Garra rufa from haplotype D; Gg–Garra ghorensis from the southern basin of the Dead sea; Gj–Garra Jordanica from the northern basin of the Dead sea; GnA–Garra nana from haplotype A; GnB–Garra nana from haplotype B. (A) Map showing the sampling sites in a regional context. Different Garra haplotypes were denoted by differently colored circles in each sampling site on the regional map. Map was reprinted from DMY + NordNordWest (https://commons.wikimedia.org/wiki/File:Israel_by_israeli_law_adm_location_map.svg), under a CC BY-SA 4.0 license, original copyright 2017. (B) A neighbor-joining unrooted tree constructed for haplotypes obtained from samples collected in this study alongside haplotypes mined from BOLD. Next to the edges are haplotype acronyms as indicated above and next to bifurcations are percent bootstrap support values. Haplotype text colors in the tree match circle colors in the map (A).