Estuaries across the globe have been subject to extensive abiotic and biotic changes and are often monitored to track trends in species abundance. The San Francisco Estuary is a novel ecosystem that has been deeply altered by anthropogenic factors, resulting in fish declines over the past 100 years. To track these species declines, a patchwork of monitoring programs has operated regular fish surveys dating back to the late 1950s. While most of these surveys are designed to track population-scale changes in fish abundance, they are methodologically distinct, with different target species, varying spatial coverage and sample frequency, and differing gear types. To remediate for individual survey limitations, we modeled pelagic fish distributions with integrated data from many sampling programs. We fit binomial generalized linear mixed models with spatial and spatiotemporal random effects to map annual trends in the distribution of detection probabilities of striped bass, Delta smelt, longfin smelt, threadfin shad, and American shad for the years 1980 to 2017. Detection probabilities decreased dramatically for these fishes in the Central and South Delta, especially after the year 2000. In contrast, Suisun Marsh, one of the largest tidal marshes on the west coast of the United States, acted as a refuge habitat with reduced levels of decline or even increased detection probabilities for some species. Our modeling approach demonstrates the power of utilizing disparate datasets to identify regional trends in the distribution of estuarine fishes.