Invasive species present one of the greatest threats to the conservation of biodiversity. When invasives hybridize with endangered native taxa, they introduce novel challenges ranging from the identification of hybrids in the field, to hybrid vigor and the erosion of species identity as genotypes are lost. Across a large swath of central California, a hybrid swarm consisting of admixed endangered California tiger salamanders (CTS, Ambystoma californiense) and introduced barred tiger salamander (Ambystoma mavortium) has replaced native populations, threatening CTS with genomic extinction. Here we employ a large-scale, genomically-informed field ecological experiment to test whether habitat restoration can reinstate natural selection favoring native salamander genotypes. We constructed 14 large, semi-natural ponds and manipulated their hydroperiods to evaluate larval survival and mass at metamorphosis. Consistent with earlier work, we found overwhelming evidence of hybrid superiority which persisted across all hydroperiod treatments. Short duration ponds substantially reduced the mass and survival probability of both native and hybrid larvae, likely exerting strong selective pressure in the wild. We identified 86 candidate genes, representing 1.8% of 4,723 screened loci, that significantly responded to this hydroperiod-driven selection. In contrast to previous mesocosm-based studies, native CTS never exhibited greater fitness than hybrids, suggesting that hydroperiod management alone will not shift selection to favor native genotypes. However, shortening pond hydroperiod may represent a cost-effective strategy to limit the overall productivity of ponds with non-native genotypes, complimenting additional strategies such as targeted hybrid removal. At a broader level, our experimental approach leverages extensive ecological knowledge, modern genomic tools, and a naturalistic, in situ replicated design to critically evaluate and expand the potential toolkit that managers can use to address this, and other recalcitrant biological invasions. We believe that this strategy may be an important tool for managing the growing number of complex invasion scenarios threatening global biodiversity.