Shifts in biodiversity and physical structure of seagrass beds across 5 decades at Carriacou, Grenadines

near complete loss of these protective structures by 2016 when we observed these reefs (S4 Fig). In a rapid assessment of the coral reefs of Carriacou in 2005, Kramer and Dowksza [37] noted extensive physical damage caused by hurricanes Ivan (2004) and Emily (2005) to many reefs along the eastern side of the island, and coral bleaching all around the island in fall 2005. These occurrences in Carriacou are consistent with the general pattern of decline in the struc- tural complexity of Caribbean coral reefs described by Alvarez-Filip et al. [33]: a steep decline in rugosity between 1969 and 1985, followed by relative stasis until 1998, and then a resump- tion in decline. These authors attribute the early decline to disease-mediated mortality of the structurally dominant Acroporoids and the latest decline to bleaching events affecting the slower growing massive corals such as Montastrea spp that remained as the major reef builders after the Acroporoid die-off. In Grenville Bay, Grenada, Reguero et al. [93] used wave energy models to explore erosional effects of increases in sea level and wave height between 1950 and 2008. They concluded that these factors were less important than the degradation of adjacent healthy coral reefs in explaining localized erosion within the bay. In a comparison of seagrass beds in three coral reef lagoons of the Veracruz Reef System (SW Gulf of Mexico), Terrados and Ramı´rez-Garcı´a [89] observed that seagrass cover was lowest, and fragmentation and blowout number highest,

by a reef that had been used as a source of building materials. They attribute this to increasing wave exposure in the lagoon through a direct anthropogenic reef-flattening effect. The stability of seagrass ecosystems, and their associated ecosystem services, including accu- mulation and long-term storage of carbon in associated sediments, has been highlighted as a key unknown in blue carbon science [16, 94]. Our 47-year long dataset suggests persistence of seagrass beds in Carriacou, although observations of increased blowouts and shifts in dominant seagrass species may indicate possible threat to carbon storage capacity. The availability of new accommodation space with sea level rise recently has been recognized as a potential mechanism of expansion of coastal wetlands [28] and blue carbon stores [19, 29]. Limiting development of new anthropogenic barriers to inland wetland migration or removing existing barriers could enable the persistence of coastal wetlands and associated benefits [28, 95]. Such barriers did not exist at the Carriacou HWE sites. Yet, our observations suggest that the potential of landward migration of seagrass to increase cover and enhance carbon stocks of reef-protected seagrass beds under rapidly rising sea levels in future [29] is unlikely to be realized at Carriacou, barring a significant reversal in the processes that have caused reef flattening. Supporting information S1 Table. Transect and station geographic information. Locations and sites (with GPS coor-

dinates) for (A) 10 transects and (B) 17 stations sampled between 1969 and 2016. The number- ing system follows that used in our paper on distribution of the invasive seagrass Halophila stipulacea in Carriacou (Scheibling et al. 2018), in which positions of transects and stations were numbered sequentially around the island. (This included 3 transects conducted in 2016 only and not shown here.) Also given are direction from shore, depth (m) at 50 m offshore, and relative wave exposure (low, LWE; high, HWE) for transects; and distance offshore (m), depth (m, chart datum) at sampling grid, habitat descriptions for stations, and the iNaturalist Observation Number for photographs of species and sites. Link directly to the Observation Number or append it to: https://inaturalist.org/observations/. View iNaturalist Project at: https://inaturalist.org/projects/seagrass-bed-flora-fauna-of-carriacou-grenada-2016. (DOCX) S2 Table. Species or species groups (denoted *, see footnotes) categorized by broad taxo- nomic group. Presence of species or species group in at least one station (1) or none (0), PLOS ONE Seagrass bed biodiversity and structure and across 5 decades PLOS ONE | https://doi.org/10.1371/journal.pone.0306897 August 1, 2024 19 / 25