Driving factors of biogeographical variation in seagrass herbivory.

Despite the crucial role of herbivory in shaping community assembly, our understanding on biogeographical patterns of herbivory on seagrasses is limited compared to that on terrestrial plants. In particular, the drivers of such patterns remain largely unexplored. Here, we used a comparative-experimental approach in Cymodocea nodosa meadows, across all possible climate types within the seagrass distribution, 2000 km and 13° of latitude in two ocean basins, to investigate biogeographical variation in seagrass herbivory intensity and their drivers during July 2014. Particularly, the density and richness of herbivores and their food resources, seagrass size, carbon and nitrogen content, as well as latitude, sea surface temperature, salinity, chlorophyll, and sediment grain size, were tested as potential drivers. We found that shallow meadows can be subjected to intense herbivory, with variation in herbivory largely explained by fish density, seagrass size, and annual sea temperature range. The herbivorous fish density was the most important determinant of such variation, with the dominant seagrass consumer, the fish Sarpa salpa, absent at meadows from regions with low herbivory. In temperate regions where herbivorous fish are present, annual temperature ranges drive an intense summer herbivory, which is likely mediated not only by increased herbivore metabolic demands at higher temperatures, but also by higher fish densities. Invertebrate grazing (mainly by sea urchins, isopods, amphipods, and/or gastropods) was the dominant leaf herbivory in some temperate meadows, with grazing variation mainly influenced by seagrass shoot size. At the subtropical region (under reduced annual temperature range), lower shoot densities and seagrass nitrogen contents contributed to explain the almost null herbivory. We evidenced the combined influence of drivers acting at geographic (region) and local (meadow) scales, the understanding of which is critical for a clear prediction of variation in seagrass herbivory intensity across biogeographical regions.