Select benthic invertebrate and sediment data from 1985 to 2023 for mapping benthic biodiversity to facilitate future sustainable development

This dataset underpins the study Mapping benthic biodiversity to facilitate future sustainable development, which used benthic macrofaunal records and sediment particle size analysis (PSA) to generate high-resolution maps of biodiversity across the continental shelf of the North-East Atlantic. The dataset comprises 22,793 standardised grab and core samples collected between 1985 and 2023, with taxonomic names harmonised via the World Register of Marine Species (WoRMS). It includes species abundance data, diversity metrics (alpha-, beta- and gamma-diversity), and biodiversity cluster classifications derived through Random Forest modelling. The primary purpose of this dataset is to inform marine spatial planning and environmental licensing, especially in the context of offshore wind development and other seabed activities.

Benthic macrofaunal records were compiled from the OneBenthic database, covering samples collected between 1985 and 2023, with most data from 2000 onwards. Only comparable grab and core samples (0.1 m², 1 millimetre mesh sieve) were retained, and taxonomic names were standardised using the World Register of Marine Species (WoRMS). Samples from areas with localised disturbance (such as dredge spoil disposal or aggregate extraction) were excluded, and those within 50 metres of one another were filtered to reduce spatial autocorrelation.

Biodiversity metrics (species richness, Shannon and Simpson diversity, abundance, and alpha-, beta-, gamma-diversity) were calculated and standardised for sampling effort. Cluster classifications and Random Forest models were applied to predict biodiversity patterns across the region, with validation through cross-validation and accuracy statistics.

Overall, the dataset is high quality and suitable for marine biodiversity mapping and planning, though users should be aware of its long temporal span and uneven sampling density across regions.

Details of how the full dataset was queried and extracted are provided in the SQL query contained within the R code.