Abstract
Assessing biodiversity and understanding how it works is a prerequisite for species conservation. The Amazon basin is one of the main biodiversity hotspots where fish are heavily exploited for ornamental purposes. The ornamental trade heavily exploits the genus Apistogramma, which is one of the most species-rich among Neotropical cichlids with 94 formally described species. This number is certainly underestimated owing to the limitations of conventional taxonomy, which is still too often based solely on morphological criteria and sometimes on few individuals. Most species of this genus have a high degree of endemism and are highly prized on the ornamental market, which could put them at risk. A few species are supposed to have extensive distributions, and in particular Apistogramma agassizii, present from the Amazon estuary up to the Ucayali and Marañon rivers in Peru. This study assessed the taxonomic status of 1,151 specimens of A. agassizii collected from 35 sites around Iquitos in the Peruvian Amazon. On the basis of molecular analyses (nuclear and mitochondrial DNA) and mate choice experiments, at least three biological species within the nominal A. agassizii were evidenced in the sampling area, which is extremely small compared with the known distribution of the species as initially described. According to the molecular calibrations, these three species would have diverged during the Plio-Pleistocene. Two of them seem to be endemic from small sub-basins, one from the Nanay River and the other from the Apayacu/Ampiyacu systems. A possible scenario that may explain the evolutionary history of these species is proposed. The conservation implications of these results on the estimation of the diversity of A. agassizii, of Apistogramma species in general, and of other Amazonian cichlids are discussed.