Effects of green turtle herbivory on seagrass ecosystems: an experimental test in the context of a global meta-analysis

Marine megaherbivores, particularly green turtles Chelonia mydas , graze on tropical seagrasses worldwide and play a substantial role in determining the structure and function of these ecosystems. To advance our understanding of the effects of green turtles on seagrass meadows, we (1) conducted an experimental exclusion experiment on turtlegrass Thalassia testudinum at 2 green turtle foraging sites in Bimini, Bahamas, and (2) performed a global meta-analysis of 587 comparisons of seagrass response variables from 32 peer-reviewed studies examining the effects of grazing by green turtles via experimental exclosure and/or simulated grazing. In our field experiment, only one turtlegrass parameter responded significantly to the exclusion of turtle grazing: across the entire 3 mo exclosure period, only seagrass density was significantly increased in the grazed plots. Across the existing literature, however, excluding green turtles has generally been found to increase seagrass aboveground biomass, shoot morphology, productivity, and metabolic rate, and not significantly affect measures of nutrient content, belowground biomass, or epiphytes. These effects were influenced by seagrass identity and grazing type but not by geographic location (i.e. latitude) or experimental duration. Our synthesis supports the general importance of green turtles in structuring seagrass ecosystems, while our experiment highlights that those effects can be site- and context-dependent, indicating the need for additional factors to be considered when assessing grazer effects on seagrass meadows.