SIBES: Long-term and large-scale monitoring of intertidal macrozoobenthos and sediment in the Dutch Wadden Sea

Abstract

The Wadden Sea is the world's largest intertidal area and a UNESCO World Heritage Site. Macrozoobenthic invertebrates perform key ecological functions within intertidal areas by regulating nutrient cycles, decomposing organic matter, and providing food for fish, birds and humans. To understand ecological processes and human impacts on biodiversity, the Synoptic Intertidal BEnthic Survey (SIBES) has sampled intertidal macrozoobenthos since 2008. On average 4,109 stations across 1,200 km² of Dutch Wadden Sea mudflats are sampled from June to October to quantify the benthic invertebrate community and sediment composition, including species abundance and biomass, and grain size and mud content. The dataset published now contains 51,851 sampled stations with 3,034,760 individuals of 177 species. This paper details data collection, validation and processing methods. SIBES is ongoing and data will be updated yearly. In sharing these data, we hope to enhance collaborations and understanding of the impact of various pressures on macrozoobenthic invertebrates, sediment composition, food webs, the ecosystem, and biodiversity in the Wadden Sea and other intertidal habitats.

Fig. 1

SIBES design with sampling stations and workflow of data collection and processing. The map of the sampling stations shows all stations that have been sampled at least once, with grid (in blue) and random sampling stations (in yellow). The inset shows a map of the detailed design. Tidal basins are separated with black lines. The workflow shows collecting samples in the field, processing them in the laboratory, and curating the data afterwards. Photos from left to right: sampling sediment cores from the boat, sampling by foot, sieved sample, jars with collected samples, and processed samples in crucibles ready for biomass and shell weight measurements in the laboratory. Photos taken by Fred Wiering (left photo) and Kees van de Veen.

Fig. 2

Data examples of the spatial extent and coverage of SIBES. (A) Sediment data with median grain size (mgs, in micrometer) per sampling station in 2010. (B) Biodiversity as the number of species per sampling station in 2021. (C) Total biomass density (g AFDM m⁻²) per sampling station in 2015.

Fig. 3

Mean biomass and abundance of benthic invertebrates in the Dutch Wadden Sea. (A) Mean annual biomass (g AFDM m⁻²) calculated across 14 years for the 15 most abundant species and all other species summed, ranked from most abundant to least abundant. (B) Mean annual numerical abundance (# m⁻²) calculated across 14 years for the 15 most abundant species and all other species summed. (C) Total biomass (g AFDM m⁻²) per year for the four most abundant species. (D) Total abundance (# m⁻²) per year for the four most abundant species. Error bars in (A) and (B) denote standard errors of the mean among years. For all species, the colors match between panels.

Fig. 4

Example of how spatial autocorrelation estimates can be used to interpolate biomass between sampling stations. The spatial sampling design of SIBES with grid and random stations combined allows for accurately estimating autocorrelation. (A) The correlogram showing the spatial autocorrelation for Mudshrimp (Corophiidae) at the Balgzand area in 2011. Correlograms can be used to interpolate the abundance of a species between sampling stations as shown in (B) where Mudshrimp densities (AFDM m⁻²) were interpolated (mapped) at a 25 × 25 m grid. Actual sampling stations are shown in solid black circles.

Fig. 5

Timing of sampling campaigns for 2008 through 2021. Each green bar represents a sampling period. The sampling period varies somewhat among years due to logistical constraints (including weather) and occurred somewhat later in 2008 and 2021.

Fig. 6

Example of a species-specific allometric relationship used for predicting an individual’s Ash Free Dry Mass (AFDM, g) from (estimated) length (mm). The solid green line is the fitted LOESS-model that is used for predicting AFDM from length. The LOESS curve was fitted using the ‘loess’ function in R, with a span of 0.6. The model’s residual standard error was 0.15. Open circles are individual measurements (n = 58,989) and grey solid dots are outliers (n = 560; corresponding to 0.9% of the data points) that fall outside of twice the inter quartile range.