Results

We compared two DNA barcode sequences of T. starostini with sequences of a total of 159 other nemacheilid loach specimens belonging to at least (due to some undescribed/unidentified species) 86 species and including all species of Paracobitis except P. ghazniensis from the Helmand drainage in Afghanistan and P. basharensis from Iran. We also include all nemacheilid loach genera recognised from the Middle East (Eidinemacheilus, Oxynoemacheilus, Paracobitis, Paraschistura, Sasanidus, Seminemacheilus, and Turcinoemacheilus), Europe (Barbatula, Oxynoemacheilus), Africa (Afronemacheilus) and western India (Aborichthys, Acanthocobitis, Acoura, Indoreonectes, Mesonoemacheilus, Nemachilichthys, Paracanthocobitis, and Schistura). From Central Asia, only Dzihunia and Triplophysa could be included as no materials were available from Iskandaria and “ Nemacheilus oxianus ”. All applied tree-reconstruction methods (ML, MP, NJ) place Troglocobitis starostini within Paracobitis (Fig 1). With a minimum interspecific K2P distance of 7.19% P. persa was the closest hit in our dataset, one of the ten species of Paracobitis included. Within the latter genus, we observed an average interspecific K2P distance of 5.43% (range 2.78–13.78%). Our molecular data do not support the view that Troglocobitis represent its own genus and we reject the hypothesis by Parin (1983) and Prokofiev (2009). As these authors suggest, it is related to Paracobitis and even nested within Paracobitis. Therefore, Troglocobitis is treated as a synonym of Paracobitis and T. starostini, as a species of Paracobitis, treated as Paracobitis starostini.