Hybridization often occurs before speciation is completed. What would be the fate of hybrid populations? Hybridization may lead to the emergence of a panmictic population (speciation reversal), hybrid speciation, co-existence of two populations, extinction of one species, or somewhere between them. Although there are many cases of hybridization, few studies have monitored long-term changes in genetic structures at the genome-wide levels. Here, we investigated the genomic change over nine years of hybrids between two stickleback fish species, freshwater Gasterosteus aculeatus and anadromous G. nipponicus, in habitats created by the 2011 Tohoku earthquake and tsunamis. Our genomic analysis showed that hybridization likely occurred in 2011, followed by backcrossing toward freshwater G. aculeatus. Lower introgression rates and faster purging rates of G. nipponicus alleles were detected on a neo-sex chromosome, which contributes to behavioral isolation, compared to autosomes. These results suggest that G. nipponicus alleles have been purged from hybrid populations, even though the genome was once admixed. We suggest that hybrid incompatibility and/or natural selection against G. nipponicus alleles may contribute to such purging. Further long-term analysis should be conducted on other hybrid populations, and comparison among different systems will help to predict the fate of hybrid populations.