Darwin theory of natural selection provides two seemingly contradictory hypotheses for explaining the success of introduced species: 1) the preadaptation hypothesis posits that introduced species that are closely related to native species will be more likely to succeed than distantly related invaders because they already possess relevant characteristics; 2) the limiting similarity hypothesis posits that invaders that are more similar to resident species will be less likely to succeed due to competitive exclusion. Previous studies assessing this conundrum show mixed results, possibly stemming from variation in study spatial scales and lack of both functional and phylogenetic information. We used species abundances compiled in a 33-year grassland successional survey based at Cedar Creek Ecosystem Science Reserve (USA) to assess the support for the pre-adaptation and limiting similarity hypotheses at two different spatial scales (neighbourhood scale of 0.5m2, site scale of ~40m2). We combined compositional surveys of 303 vascular plant taxa (256 native, 47 introduced) taken across 2700 plots in a chronosequence of abandonment from agriculture with species functional dissimilarities, phylogenetic distances, environmental covariates and information on species origin. Our results consistently supported the pre-adaptation hypothesis at the site scale but diverged at neighbourhood scale, with functional dissimilarity supporting the limiting similarity hypothesis and phylogenetic distance supporting the preadaptation hypothesis. Introduced species with low leaf dry matter content (LDMC), low height and high seed mass tended to be most abundant than rest of species, while relationships between species abundance and specific leaf area (SLA) varied with scale. Introduced species were more abundant than natives at higher concentrations of soil N but were less abundant than natives over time. Our study highlights the importance of environmental filtering on grassland community assembly at the scale of a site (here 40 m2, a spatial resolution that is usually considered local). This influence of environmental filtering might mask effects of limiting similarity at small-local scales. Our results demonstrate the importance of accounting for both phylogenetic and functional dissimilarity when examining the complex interaction between species biogeographic origin, functional strategies and evolutionary history as considering one alone can lead to different conclusions. ### Competing Interest Statement The authors have declared no competing interest.