Abstract
Human disturbance modifies selection regimes, depressing native species fitness and enabling the establishment of non-indigenous species with suitable traits. A major impediment to test the effect of disturbance on invasion success is the lack of long-term data on the history of invasions. Here, we overcome this problem and reconstruct the effect of disturbance on the invasion of the bivalve Anadara transversa from sediment cores in the Adriatic Sea. We show that (1) the onset of major eutrophication in the 1970s shifted communities towards species tolerating hypoxia, and (2) A. transversa was introduced in the 1970s but failed to reach reproductive size until the late 1990s because of metal contamination, resulting in an establishment and detection lag of ~25 years. Subfossil assemblages enabled us to (1) disentangle the distinct stages of invasion, (2) quantify time-lags and (3) finely reconstruct the interaction between environmental factors and the invasion process, showing that while disturbance does promote invasions, a synergism of multiple disturbances can shift selection regimes beyond tolerance limits and induce significant time lags in establishment. The quantification of these time lags enabled us to reject the hypothesis that aquaculture was an initial vector of introduction, making shipping the most probable source.