Database generated through a systematic review comparing seagrass and adjacent unvegetated sediments, focusing on carbon storage, pollution mitigation, and benthic biodiversity enhancement.

Database associated to the scientific article: Global meta-analysis reveals variability in seagrass carbon storage, pollution mitigation and benthic biodiversity enhancement provisioning. This database is a compilation of published data on variables proxy to assess the effect of the presence of seagrasses enhances three key services: (1) organic carbon accumulation rates (CAR), (2) water purification through metals, microplastics and nutrients burial (WAT), and (3) infaunal benthic biodiversity support (BIO). Systematic review methods: Data were compiled through a systematic review following the PRISMA protocol (Page et al., 2021) and using the ISI Web of Knowledge database for peer-reviewed reports in English (https://www.webofknowledge.com, accessed 20/07/2021). The search terms and logical operators used were as follows: “TS = ((soil OR sediment) AND (seagrass*))”. TS indicates that the search retrieves records where the specified terms appear in the Title, Abstract, or Keywords. The results of the search yielded 4414 records from 1982 to 2021. The records obtained were cross-checked against publications from the authors of this review, to assess if key reports were retrieved. This assessment showed that the search did not properly cover biodiversity related literature. Therefore, a second search was conducted specifically targeting biodiversity data: “TS = (seagrass AND (infauna* OR benth* fauna* OR benthos))”, accessed 25/05/2022. The second search yielded 723 records. Both reviews were updated in May 2025 to screen papers published between 2022 and May 2025, yielding an additional 910 publications. The identified records were selected if they reported any of the target variables in both seagrass meadows and bare sediment (Table 1). The methods section was checked, as information about control cores in bare sediment was often not included in the abstract. For each target ecosystem service, we selected proxy variables, to assess the effect of seagrass presence over ecosystem service provision, and complementary variables, to discuss differences in ecosystem services provision, based on the authors' expertise (Table 1). From each record, the data on these variables were extracted from the text, figures, and tables of each publication. When necessary, data from figures were extracted using the software WebPlotDigitizer (Rohatgi, 2017) or metaDigitise (Pick et al., 2018). At the data extraction phase, reports were further excluded if: (1) the control data was not from bare sediment (e.g. from other habitat types such as other seagrass meadows, mangrove forests, oyster beds, etc.); (2) the bare sediment was a degraded or restored meadow; and (3) the bare sediment was not adjacent to the seagrass meadow (e.g., the bare sediment was collected from another bay or estuary). Table 1. List of extracted variables. Type indicates whether the variable is a Proxy, variable, selected to assess the effect of seagrass presence over ecosystem service provision, or a Complementary, variable, used to discuss differences in ecosystem service provisioning. Variable Units Type Number of extracted data pairs Number of data pairs in final models CAR - Carbon accumulation Organic carbon (OC) content % dw Proxy 2924 260 Organic matter (OM) content % dw Proxy 547 167 OC stock g cm-2 Proxy 328 42 Carbon density g cm-3 Proxy 2354 20 OM accumulation rate g cm-2 day-1 Proxy 12 1 Mud content % dw Complementary 352 92 Dry bulk density g cm-3 Complementary 115 59 Mean grain size mm Complementary 109 34 Median grain size mm Complementary 73 28 Organic C to N ratio Complementary 17 13 Total C to N ratio mass Complementary 28 7 Total C content % dw Complementary 44 7 Total C to N ratio molar Complementary 23 6 Contribution of seagrass as an OC source % Complementary 5 1 WAT - water purification N content % dw Proxy 225 93 P content % dw Proxy 29 14 N stock g cm-2 Complementary 5 4 Microplastics content per mass particles kg-1 dw Proxy 74 27 Al concentration ppm Proxy 1 1 As concentration ppm Proxy 41 17 Cd concentration ppm Proxy 31 13 Co concentration ppm Proxy 26 8 Cr concentration ppm Proxy 46 18 Cu concentration ppm Proxy 50 24 Fe concentration ppm Proxy 49 20 Hg concentration ppm Proxy 14 6 Mg concentration ppm Proxy 4 1 Mn concentration ppm Proxy 41 23 Ni concentration ppm Proxy 42 17 Pb concentration ppm Proxy 35 17 Sr concentration ppm Proxy 1 1 V concentration ppm Proxy 19 7 Zn concentration ppm Proxy 55 24 BIO - biodiversity support Density per area ind. m-2 Proxy 452 146 Biomass per area g m-2 Proxy 268 80 Species richness # of species Proxy 364 67 Shannon-Wiener diversity index Proxy 245 54 Pielou evenness index Proxy 177 30 Simpson's diversity index Proxy 21 11 Margalef's diversity index Proxy 60 6 Density per volume ind. m-3 Proxy 3 3 Simpson evenness index (1-D) Proxy 2 2 Data were extracted as paired seagrass and control (bare sediment) samples. For each data pair we recorded the location, country, seagrass bioregion (Short et al. 2007), latitude, longitude, if the meadow was mono specific or composed of multi-species, dominant seagrass species, if data were from a whole core or from a subsample, sampling mesh size (for BIO variables), minimum and maximum sediment depth of the sample/core, and known anthropogenic impacts in the area. Furthermore, if it was possible to extract this information from the publication, we also annotated the sampling year, secondary or tertiary seagrass species, tidal range, geomorphology, and hydrodynamic exposure of the area. Mean value, sample size and standard deviation were recorded for each location, both for the seagrass meadow and the bare area. If the standard deviation was not provided it was calculated based on the standard error and sample size. In cases where multiple stations were reported without a measure of heterogeneity (e.g., standard deviation or standard error), data from all stations were combined to calculate an average and standard deviation for both the seagrass and bare areas of that study. Infaunal groups were pooled by size class into micro- (<0.2 mm), meio- (0.2-0.75 mm), and macrofauna (>0.75 mm) to better enable comparisons of faunal communities and assess whether habitat provisioning of seagrass meadows differs by size class. The size classes were assigned based on the mesh and/or sieve sizes reported in the studies. Seagrasses were also divided into functional groups (i.e. opportunistic, colonizing, and persistent, Table 2) based on the definition by Kilminster et al. (2015). From the resulting database we selected the most sampled variables as indicators of each ecosystem service: OC and OM content were selected for CAR; P, N, microplastics, and metals (pooled) content for WAT; and abundance per area, density per area, species richness, Shannon-Weiner and Pielou diversity index for BIO. The database is presented in five .csv files. Specific .csv files content: Sources.csv ● or_study_id: Original study ID ● study_id: Final study ID ● user: Researcher who extracted data froom this source. ● journal: Journal name (full). ● title: Publication title. ● year: The year of publication (YYYY). ● doi: Digital object identifier associated with the work. ● annotation: Notes taken during source screening. Database.csv ● study_id: Final study ID. Same as in Sorces.csv ● site: Site where the samples were collected. ● country: Country where the samples were collected. In English. ISO 3166 international standard. ● bioregion: Seagrass bioregion as defined in Short et al. 2007. ● lat: Positional latitude of the sample in decimal degree WGS84. General for the sample location, not for each type of habitat. Positive for N, negative for S. ● long: Positional longitude of the sample in decimal degree WGS84. General for the sample location, not for each type of habitat. Positive for E, negative for W. ● sampling_year: Year of the core sampling. Format: YYYY. ● meadow_type: Mono-specific or multi-specific seagrass meadow. ● species_1: First most abundant species. ● species_2: Second most abundant species. ● species_3: Third most abundant species. ● tidal_range: Intertidal, shallow subtidal (< 10 m), deep subtidal (≥ 10 m), or unspecified. ● geomorphology: Type of geomorphology of the site. Levels: estuary, lagoon, open nearshore, open offshore, others. ● hydrodinamic_exposure : Qualitative category of exposure to waves and currents. Two levels: protected or exposure. Select “unspecified” if not given. ● section: Whether the data extracted is for the whole core or for a layer inside it. ● core_depth: Yes: 1; No: 0 ● min_depth: Minimum sediment depth for the core or layer (cm) = top cm. ● max_depth: Maximum sediment depth for the core or layer (cm) = bottom cm. ● car_time: Time range used for the estimation of the burial rate (years). ● bg_y_n: Is the belowground biomass included in the soil analysis. Yes: 1; No: 0 ● origin_method: Method used to estimate the origin (source) of the organic matter. ● organism_group: Group of organisms studied in the biodiversity assessment. Compulsory if the selected variable is from category BIO. ● variable : List of variables, with units, given by category: CAR, carbon sequestration, WAT, water purification, BIO, biodiversity support. ● seagrass_core_id: Core ID for the seagrass. ● seagrass_val: Mean value for that variable in the habitat SEAGRASS. ● seagrass_se: Standard error of the mean in the habitat SEAGRASS. ● seagrass_sd: Standard deviation of the mean in the habitat SEAGRASS. ● seagrass_n: Sample size associated in the habitat SEAGRASS. ● bare_core_id: Core ID for bare sediment. ● bare_val: Mean value for that variable in the habitat BARE SEDIMENT. ● bare_se : Standard error of the mean in the habitat BARE SEDIMENT. ● bare_sd: Standard deviation of the mean in the habitat BARE SEDIMENT. ● bare_n: Sample size associated in the habitat BARE SEDIMENT. ● antrpogenic_impacts: area studied was under any antropogenic impact. ● mesh_size: mesh size used to sieve benthic organisms. Compulsory if the selected variable is from category BIO. ● Annotation: Notes taken during data extraction.