BackgroundThe Hajar Mountains of Oman and the United Arab Emirates (UAE) is the highest mountain range in Eastern Arabia. As a result of their old geological origin, geographical isolation, complex topography and local climate, these mountains provide an important refuge for endemic and relict species of plants and animals with strong Indo-Iranian affinities. Among vertebrates, the rock climbing nocturnal geckos of the genus Asaccus represent the genus with the highest number of endemic species in the Hajar Mountains. Recent taxonomic studies on the Zagros populations of Asaccus have shown that this genus is much richer than it was previously thought and preliminary morphological and molecular data suggest that its diversity in Arabia may also be underestimated.MethodsA total of 83 specimens originally classified as Asaccus caudivolvulus (including specimens of the two new species described herein), six other Asaccus species from the Hajar and the Zagros Mountains and two representatives of the genus Haemodracon were sequenced for up to 2,311 base pairs including the mitochondrial 12S and cytb and the nuclear c-mos, MC1R and ACM4 genes. Phylogenetic relationships were inferred using both Bayesian and maximum-likelihood approaches and the former method was also used to calibrate the phylogenetic tree. Haplotype networks and phylogenetic trees were inferred from the phased nuclear genes only. Sixty-one alcohol-preserved adult specimens originally classified as Asaccus caudivolvulus from the northern Hajar Mountains were examined for 13 morphometric and the five meristic variables using multivariate methods and were also used to diagnose and describe the two new species.ResultsThe results of the molecular and morphological analyses indicate that the species originally classified as Asaccus caudivolvulus is, in fact, an assemblage of three different species that started diversifying during the Mid-Miocene. The molecular phylogenies consistently recovered the Hajar endemic A. montanus as sister taxon to all the other Asaccus species included in the analyses, rendering the Arabian species of Asaccus polyphyletic.DiscussionUsing this integrative approach we have uncovered a very old diversification event that has resulted in a case of microendemicity, where three morphologically and ecologically similar medium-sized lizard species coexist in a very short and narrow mountain stretch. Asaccus caudivolvulus is restricted to a small coastal area of the UAE and at risk from heavy development, while the two new species described herein are widely distributed across the northern tip of the Hajar Mountains and seem to segregate in altitude when found in close proximity in the Musandam Peninsula (Oman). Similarly to other integrative analyses of Hajar reptiles, this study highlights the high level of diversity and endemicity of this arid mountain range, underscoring its status as one of the top hotspots of reptile diversity in Arabia.

The Hajar Mountains of south-eastern Arabia form an isolated massif surrounded by the sea to the east and by a large desert to the west. As a result of their old geological origin, geographical isolation, complex topography and local climate, these mountains provide an important refuge for endemic and relict species of plants and animals. With 19 species restricted to the Hajar Mountains, reptiles are the vertebrate group with the highest level of endemicity, becoming an excellent model for understanding the patterns and processes that generate and shape diversity in this arid mountain range. The geckos of the Ptyodactylus hasselquistii species complex are the largest geckos in Arabia and are found widely distributed across the Arabian Mountains, constituting a very important component of the reptile mountain fauna. Preliminary analyses suggested that their diversity in the Hajar Mountains may be higher than expected and that their systematics should be revised. In order to tackle these questions, we inferred a nearly complete calibrated phylogeny of the genus Ptyodactylus to identify the origin of the Hajar Mountains lineages using information from two mitochondrial and four nuclear genes. Genetic variability within the Hajar Mountains was further investigated using 68 specimens of Ptyodactylus from 46 localities distributed across the entire mountain range and sequenced for the same genes as above. The molecular phylogenies and morphological analyses as well as niche comparisons indicate the presence of two very old sister cryptic species living in allopatry: one restricted to the extreme northern Hajar Mountains and described as a new species herein; the other distributed across the rest of the Hajar Mountains that can be confidently assigned to the species P. orlovi. Similar to recent findings in the geckos of the genus Asaccus, the results of the present study uncover more hidden diversity in the northern Hajar Mountains and stress once again the importance of this unique mountain range as a hot spot of biodiversity and a priority focal point for reptile conservation in Arabia.

The Persian Horned Viper (Pseudocerastes persicus) is distributed from northeast Iraq through the Iranian Plateau to western Pakistan with isolated populations in the Hajar Mountains of south-eastern Arabia. Like the other members of the genus Pseudocerastes, P. persicus is a sit-and-wait ambush feeder with low vagility, a characteristic that often results in high levels of population differentiation. In order to clarify the level of genetic variability, phylogenetic relationships, and biogeography of the Arabian populations of P. persicus we sequenced 597 base pairs of the mitochondrial cytochrome b of four individuals from the Hajar Mountains in south-eastern Arabia and inferred their phylogenetic relationships including 10 samples of P. persicus from Iran and Pakistan, four P. urarachnoides and one P. fieldi downloaded from GenBank. The four Arabian samples are genetically very similar in the gene fragment analysed and are phylogenetically very closely related to populations of P. persicus from coastal south Iran. Biogeographically, it appears that colonisation of the Hajar Mountains by P. persicus took place from Iran very recently, most probably during the last glaciation, when most of the Persian Gulf was above sea level and did not represent a barrier for dispersal.

The spawning season of the deep sea smelt of Japan Sea was found to be very long, extending to several months. Hence it is believed that two or more races could be spawning one after the other. On the basis of otolith patterns this fish population could be separated into two major groups. The developmental pattern of the otolith of the deep sea smelt has been shown in a model else where. Linear discriminant function analysis of seven meristic characters showed a distinct variation between the groups. Tests showed that the yearly variation of the meristic characters were not significant.Deep sea smelt, Glossonodon semifasciatus, as the name implied is a bottom feeding, slender, long fish living at a depth of about 100 to 150 meters. It is an important con tributor to the Danish seine fishery of Japan Sea. The fishing season of this fish falls on its spawning season, which extends from September to March. The spawning season being so long it is believed that two or more races would be spawning one after the other. This behavior has been observed in herring populations by several workers. On the basis of otolith pattern deep sea smelt population could be separated into two major groups. According to the model constructed by SINODA and JAYASINGHE10) hypothesizing the developmental pattern of the otolith of this fish, early spawners will have bigger hya line nucleii and narrower first growth zones, while late spawners may have smaller nucleii and wider first growth zones. The growth rates of the two groups showed a consider able difference. The growth constants calculated from VON BERTALANFFY'S growth curves for the two groups gave 0.5137 and 0.3682 year-1 for winter-early spring spawners and autumn spawners respectively. This may be due to varied metabolic rates under different environmental conditions or due to feeding competition in the community.The growth curves of the two groups shown in Fig. 1, seems to give good envidence in confirmation of the assumptions on which the model represents the otolith formation.There is a shift of about five months in the origin of growth curve of the autumn-early winter spawners from that of the spring spawners. This clearly shows that birth times of the two groups are different to each other. As a method of separating these two populations discrimi nant analysis was employed. Materials and MethodsSamples of both juvenile and adult deep sea smelt were collected from the commercial