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

(T12) was in an area that was part of a fringing seagrass bed in 1969 and 1994 but consisted of patches of seagrass in 2016. Thalassia testudinum was present at all 17 stations in all years and mean frequency (n = 12 quadrats per station) declined marginally (from 100 to 93%) between 1969 and 2016 (Fig 3A). Syringodium filiforme was present at 6 stations in 1969 and at 10 stations in both 1994 and 2016. Mean frequency of S. filiforme increased from 1969 (35%) to 1994 (52%) and 2016 (50%), although these changes were not statistically significant (F2,48, P = 0.548). Halophila sti- pulacea, accounted for the lowest recorded mean frequency (7.8%), occurring at only 2 of 17 stations in 2016. Halodule wrightii was not recorded in any frequency observations. Mean leaf biomass of T. testudinum per area of bottom declined by a third (from 841 to 567 wet g m-2) between 1969 and 2016 (Fig 3B). In contrast, mean leaf biomass of S. filiforme increased nearly 3-fold during this time (from 159 to 448 wet g m-2). H. stipulacea first appeared in 2016 at 2 of the 17 stations, where it represented 59% and 5% of the local biomass but <1% (4 wet g m-2) of total mean biomass for all stations. H. wrightii was absent through- out, except at one station in 2016 where biomass was minimal (5.3 g m-2; not shown in Fig 3B). Mean total seagrass biomass changed little across the 3 sampling years (1023, 943, 1001 wet g m-2 respectively). These temporal changes suggest a shift in relative biomass of the two most abundant species, with similar declines and increases for T. testudinum and S. fili-

forme respectively. However, these changes in biomass are not statistically significant for either species (1-way ANOVA, square root transformation: T. testudinum, F2,48 = 1.63, P = 0.206; S. filiforme, F2,48 = 0.863, P = 0.448). Fig 2. Substratum cover in seagrass beds. Mean (+ SE) % cover of transects in 6 HWE sites 4 LWE sites in 1969 and 2016, and 3 HWE sites and 1 LWE site in 1994 (top row). Cover on individual transects at HWE and LWE sites in each year (subsequent 3 rows). Height of bars indicates total cover of seagrass bed; remainder is cover of seagrass-free inshore zone (SFI). The large component “Other” in 2016 consists mainly of a mixture of Thalassia testudinum, Syringodium filiforme, and Halodule wrightii on Transect T9. https://doi.org/10.1371/journal.pone.0306897.g002 Fig 3. Abundance of seagrass species/groups at stations (n = 17) in 1969, 1994 and 2016. Mean (+ SE) of a) frequency of occurrence (%) and b) biomass (kg m-2) of Thalassia testudinum, Syringodium filiforme and Halophila stipulacea. (Halodule wrightii is not included as it occurred at very low frequencies and biomass at stations). https://doi.org/10.1371/journal.pone.0306897.g003 PLOS ONE Seagrass bed biodiversity and structure and across 5 decades PLOS ONE | https://doi.org/10.1371/journal.pone.0306897 August 1, 2024 9 / 25