Mangroves are particularly extensive on sheltered, macrotidal, muddy tropical coastlines, but also occur in association with coral reefs. Reefs attenuate wave energy, in some locations enabling the accretion of fine calcareous sediments which in turn favour establishment of seagrasses and mangroves. Knowledge of the distribution and ecology of both reefs and mangroves increased in the first half of the 20th century. J Alfred Steers participated in the Great Barrier Reef Expedition in 1928, and developed an interest in the geomorphological processes by which islands had formed in this setting. It became clear that many mangrove forests showed a zonation of species and some researchers inferred successional changes, even implying that reefs might transition through a mangrove stage, ultimately forming land. Valentine Chapman studied the ecology of mangroves, and Steers and Chapman described West Indian mangrove islands in the 1940s during the University of Cambridge expedition to Jamaica. These studies provided the background for David Stoddart’s participation in the Cambridge Expedition to British Honduras and his PhD examination of three Caribbean atolls. One of these, Turneffe Islands, is an atypical atoll, with mangroves occupying more than 25% of the reef platform. In contrast to this atoll-like island, there are other island types on the Belize barrier reef, including similar extensive mangrove islands called ‘ranges’. Stoddart compared the more complex of these, comprising shingle ridges and sand cays around which the mangroves occur, with the ‘low wooded islands’ on the Great Barrier Reef. Many of those that he had mapped were devastated by Hurricane Hattie in 1961, and the opportunity to revisit these, and document the storm damage and subsequent recovery, led to much important evidence for the role of storms in the long-term evolution of both reefs and reef islands. In 1973, Stoddart led the Royal Society and Universities of Queensland Expedition to the northern Great Barrier Reef. During this extensive period of fieldwork, Stoddart and colleagues remapped Low Isles in detail, the best-known of the low wooded islands, and many other islands in less detail. The contrast between islands on the Belize coast with those on the Great Barrier Reef provided many insights into the geomorphology of both reefs and the vegetation of islands that form on them. The physiography of the islands provided many clues to the way they had formed and the processes operating. It became clear to Stoddart that there has been a different sea-level history in these two locations. He was able to observe and infer the response of key habitats to tropical cyclones (hurricanes). In many cases island physiography differs between individual islands, reflecting a broader suite of factors such as inherited topography and morphodynamic feedbacks; in other cases, mangrove establishment appears to have been opportunistic. Subsequent studies have elucidated the stratigraphy and geochronology in some locations; in many cases confirming the hypotheses that Stoddart proposed. His insightful descriptions and meticulous fieldwork have provided a wealth of observational data that have inspired subsequent studies and that will continue to generate alternative hypotheses that future environmental scientists can endeavour to test.