Coral reefs are currently threatened by both natural and human-driven disturbances, and climate change scenarios suggest that the health of future reefs is uncertain. To better inform management and preservation of coral reef ecosystems, it is critical that we understand the mechanisms that support recovery of coral habitat after coral mortality events. Macroalgae can inhibit coral recovery by preventing recruitment and growth of corals after a disturbance, leading to a shift in the dominant benthic community from coral to macroalgae. These macroalgal habitats do not support the same diversity of fish and other species as healthy coral reefs, and are thus undesirable on large scales. This dissertation addresses some processes that are important in the establishment and persistence of these macroalgal phase shifts. Through an enhanced understanding of macroalgal persistence, we may be able to identify (and protect) key species and processes that are critical to promoting coral reef recovery. I used the macroalga Turbinaria ornata as a model species to examine the potential for herbivore-algal feedbacks that can promote persistent macroalgal phase shifts. This species is perennial and patches appear to persist inter-annually once established. In areas where macroalgae is present, Turbinaria largely dominates the benthic community. Results of observational studies of tagged Turbinaria patches suggest that Turbinaria populations exhibit positive population growth (both number of adult thalli and patch area increased over a 2 year period). Several factors including high turnover rates of thalli, year-round reproduction, and a buffer of immature individuals that may act as a “seed bank” following a disturbance may contribute to the expansion of this species in the lagoons of Moorea. In addition, I used field experiments to assess the potential for associational refuge to provide a mechanism for Turbinaria persistence in the lagoons of Moorea. Results from these experiments suggest that an associational refuge may facilitate persistence of Turbinaria, as predation on small vulnerable individuals (< 2 cm in length) is reduced when associated with larger, less palatable adults. However, the significance of this refuge depends on the local abundance of browsing herbivores that can consume macroalgae. An associational refuge that results in higher recruit survival in the presence of herbivory could represent a self-reinforcement mechanism that could facilitate the establishment and multigenerational persistence of Turbinaria patches on coral reefs. To determine the effects of herbivore pressure on the development and persistence of a macroalgal community, I examined the effects of herbivores on mature (dominated by the late stage macroalga Turbinaria ornata) and early stage (dominated by turf algae) algal communities. Results suggest that the effects of herbivores depend upon the initial conditions of the benthic community, and this may have implications for management practices for disturbed, degraded and healthy reefs. Herbivores, particularly browsing fish that can consume mature macroalgae, appear to be limited in their capacity to remove established macroalgae. These critical species appear to be present in low abundance in some habitats, and are targeted as part of a local fishery. In contrast, other herbivores (grazers that consume turf algae and immature macroalgae) appear to readily limit the development of macroalgal communities. I also found that herbivores differentially influence the richness of macroalgal communities depending on the initial macroalgal community. Specifically, an initial presence of the mature macroalga, Turbinaria ornata, enhanced richness of macroalgal assemblage even as herbivore pressure increased. At the same level of herbivore pressure, communities with Turbinaria present had greater richness than communities that initially were dominated by turf algae, even after two years. Results from this experiment provide some of the first empirical tests of hysteresis (path dependency) in reef recovery, and reinforce the importance of different herbivore functional groups at different points along the recovery trajectory. Overall, this research highlights characteristics of macroalgae and interactions between macroalgal and herbivore communities that can drive persistent macroalgal phase shifts in coral reef ecosystems. This information may be used to develop sustainable management and conservation strategies to preserve these valuable ecosystems and the services that they provide for local communities.