RESUMO
Insular biodiversity hotspots of Southeast Asia are remarkable for their biodiverse faunas. With a marine larval phase lasting up to several months, the freshwater fish subfamily Sicydiinae has colonized most islands of these hotspots. However, Sicydiinae diversity is still poorly understood in Southeast Asia. With the objective to estimate intraspecific genetic diversity and infer past demography, we conducted the molecular inventory of Sicydiinae species in Sundaland and Wallacea using 652 bp of the mitochondrial cytochrome oxidase I gene, species delimitation methods and Bayesian Skyline plot reconstructions. In total, 24 Molecular Operational Taxonomic Units are delimited among the 603 sequences belonging to 27 species and five genera. Two cases of discordance between morphology and mitochondrial sequence are observed suggesting ongoing speciation and/or introgression in two genera. Multiple new occurrences are reported, either for a single biodiversity hotspot or both, some of which corresponding to observations of a few individuals far from the range distribution of their conspecifics. Among the ten species or species group whose intraspecific diversity was examined, high levels of genetic diversity and past population expansion are revealed by Tajima's D tests and Bayesian Skyline Plot reconstructions. Together these results indicate that long-distance dispersal is common and suggest that most endemic species originated through founder events followed by population expansion. Patterns of sexual dimorphism and males' coloration among diverging species pair seem to point to sexual selection as an important mechanism contributing to speciation in the Sicydiinae of Sundaland and Wallacea.
RESUMO
The Coral Triangle (CT), a region spanning across Indonesia and Philippines, is home to about 4,350 marine fish species and is among the world's most emblematic regions in terms of conservation. Threatened by overfishing and oceans warming, the CT fisheries have faced drastic declines over the last decades. Usually monitored through a biomass-based approach, fisheries trends have rarely been characterized at the species level due to the high number of taxa involved and the difficulty to accurately and routinely identify individuals to the species level. Biomass, however, is a poor proxy of species richness, and automated methods of species identification are required to move beyond biomass-based approaches. Recent meta-analyses have demonstrated that species richness peaks at intermediary levels of biomass. Consequently, preserving biomass is not equal to preserving biodiversity. We present the results of a survey to estimate the shore fish diversity retailed at the harbor of Ambon Island, an island located at the center of the CT that display exceptionally high biomass despite high levels of threat, while building a DNA barcode reference library of CT shore fishes targeted by artisanal fisheries. We sampled 1,187 specimens and successfully barcoded 696 of the 760 selected specimens that represent 202 species. Our results show that DNA barcodes were effective in capturing species boundaries for 96% of the species examined, which opens new perspectives for the routine monitoring of the CT fisheries.