In the present study we report a spatial autocorrelation analysis of molecular data obtained for Arapaima gigas, and the implication of this study for conservation and management. Arapaima is an important, but critically over-exploited giant food fish of the Amazonian várzea. Analysis of 14 variable microsatellite loci and 2,347 bp of mtDNA from 126 individuals sampled in seven localities within the Amazon basin suggests that Arapaima forms a continuous population with extensive genetic exchange among localities. Weak effect of isolation-by-distance is observed in microsatellite data, but not in mtDNA data. Spatial autocorrelation analysis of genetic and geographic data suggests that genetic exchange is significantly restricted at distances greater than 2,500 km. We recommend implementing a source-sink metapopulation management and conservation model by proposing replicate high quality várzea reserves in the upper, central, and lower Amazon basin. This conservation strategy would: 1) preserve all of the current genetic diversity of Arapaima; 2) create a set of reserves to supply immigrants for locally depleted populations; 3) preserve core várzea areas in the Amazon basin on which many other species depend. We stress that conservation strategies should not only preserve current genetic diversity, but also the evolutionary processes which have generated the observed patterns.
pirarucu; gene flow; reserves; várzea; autocorrelation; microsatellites