Many species are expanding their distributions to higher latitudes due to global warming. Understanding the mechanisms underlying these distribution shifts is critical for better understanding the impacts of climate changes. The climate envelope approach is widely used to model and predict species distribution shifts with changing climates. Biotic interactions between species, however, may also influence species distributions, and a better understanding of biotic interactions could improve predictions based solely on climate envelope models. Along the northern Gulf of Mexico coast, USA, subtropical black mangrove (Avicennia germinans) at the northern limit of its distribution grows sympatrically with temperate salt marsh plants in Florida, Louisiana, and Texas. In recent decades, freeze-free winters have led to an expansion of black mangrove into salt marshes. We examined how biotic interactions between black mangrove and salt marsh vegetation along the Texas coast varied across (i) a latitudinal gradient (associated with a winter-temperature gradient); (ii) the elevational gradient within each marsh (which creates different marsh habitats); and (iii) different life history stages of black mangroves (seedlings vs. juvenile trees). Each of these variables affected the strength or nature of biotic interactions between black mangrove and salt marsh vegetation: (i) Salt marsh vegetation facilitated black mangrove seedlings at their high-latitude distribution limit, but inhibited black mangrove seedlings at lower latitudes; (ii) mangroves performed well at intermediate elevations, but grew and survived poorly in high- and low-marsh habitats; and (iii) the effect of salt marsh vegetation on black mangroves switched from negative to neutral as black mangroves grew from seedlings into juvenile trees. These results indicate that the expansion of black mangroves is mediated by complex biotic interactions. A better understanding of the impacts of climate change on ecological communities requires incorporating context-dependent biotic interactions into species range models.