ncreasing green turtle abundance will lead to increased grazing within seagrass habitats—ecosystems that are important for carbon sequestration and storage. However, it is not well understood how carbon dynamics in these ecosystems respond to grazing and whether a response differs among meadows or locations.We measured seagrass ecosystem metabolism in grazed and ungrazed areas of Thalassia testudinum meadows with established green turtle foraging areas across the Greater Caribbean and Gulf of Mexico. We sampled meadows from five locations that differed in seagrass and environmental characteristics. Established meadows of the invasive seagrass Halophila stipulacea were also present at two of these locations, and we measured ecosystem metabolism in these meadows for comparison to grazed and ungrazed areas of the native T. testudinum.Across all individual sites, rates of net ecosystem production (NEP) ranged from 56% – 96% lower in grazed areas than ungrazed areas of T. testudinum meadows. Rates of NEP were also strongly, positively correlated with aboveground seagrass biomass across sites. While metabolic carbon capture rates were lower in grazed areas, heterotrophic respiration was not stimulated, and grazing therefore did not result in significant metabolic remineralization of carbon in these meadows. Net ecosystem production in H. stipulacea meadows was similar to rates in T. testudinum meadows at all three sites, suggesting that metabolic carbon capture may remain similar in Caribbean meadows where this invasive seagrass is replacing native species.Synthesis: Our results show there is a consistent response in metabolic carbon dynamics to green turtle grazing in T. testudinum meadows across the Greater Caribbean region. An increase in grazing will not likely stimulate remineralization of carbon as these important habitats are returned to a natural grazed state.