Benthic occurrences, habitat maps, and species traits

Introduction

Understanding the substrate and habitat preferences of marine benthic invertebrates is important for a number of reasons. It influences their functional role in the ecosystem; it can inform decisions about important habiats for conservation, potentially leading to the decelopment of biological indicators of habitat types; and it helps identify communities of co-occurring species. However, as for other biological traits, basic natural history knowledge of habitat affinities is lacking for many marine invertebrates - at least, it is not routinely or consistently stored in open databases. This makes accessing such data for large numbers of species (100s or 1000s) for regional-scale assessments extremely timeconsuming and laborious.

Data products built on EMODnet Biology data now allow us to robustly map the spatial occurrence of benthic species. Thanks to EMODnet Seabed Habitats we have a broadscale habitat classification map of European seas, which are related to biological composition in varied and often complex ways. There are as yet however no formal links between EurOBIS occurrence data and the Seabed Habitats map. Matching these two sources of data together allows us therefore to summarise the substrate types where a species is known to occur, as well as the species composition of different habitat types. It can help to refine and improve habitat maps (e.g. ‘species X occurs here so there must be rock here’); it can help in the quality control of occurrence records (e.g. ‘this location must be recorded imprecisely because species X is recorded but there is no rock here’); and it can also help to improve substrate preference information about individual species (e.g. ‘species X must also like mud because it frequently occurs there’).

The aim of this project then is to link occurrences of benthic invertebrates from the EMODnet benthic numerical abundance product to two complementary seabed habitat products: the EMODnet broadscale habitat map 2019 (Vasquez et al., 2020), and Wilson et al.'s (2018) synthetic map of the north-west European Shelf sedimentary environment. We also process and incorporate qualitative information on the habitat and sediment preferences of species from the Biological Traits Information Catalogue BIOTIC database. We present one example of application of this process (generating species-level summaries of substrate affinities), and provide all the necessary code generate the data to address the other types of question (e.g. to produce species lists for different habitat types).

Directory structure

{{EMODnet_occs_habs}}/
├── analysis
├── data/
│   ├── derived_data/
│   └── raw_data/
├── docs/
├── product/
└── scripts/

• analysis - Markdown notebooks
• data - Raw and derived data
• docs - Rendered reports
• product - Output product files
• scripts - Reusable code

Data

Raw data on individual traits from BIOTIC have been included in the data/raw_data folder. The other key raw datasets are all available online, and the scripts provded here show how they are dowloaded and imported. Processed and derived datasets that are generated by the scripts are all saved to the 'derived data' folder. These are fully described in the markdown documents in which they are generated.

Analysis

There are three workbooks included in this folder:

1. 'species and traits.Rmd' describes the processing of BIOTIC habitat traits data to generate a dataset of species with documented substrate preferences and habitat affinities.
2. 'benthic data habitat matching.Rmd' describes the process of downloading, importing, and processing the benthic numerical abudance product, the broadscale seabed habitat map, and the sediment properties map, and linking these all together to generate the three key datasets that the generation of the data products requires:

• benthic_abundance_sampling_events_seabed_habs.csv, which contains information on each distinct sampling event, including its estimated seabed habitat characteristics;
• benthic_abundances_long.csv which is a minimal and memory-efficient version of the numerical abundance product;
• benthic_taxa.csv which contains taxonomic information on all the taxa included in the numerical abundance product.

3. 'benthic species habitat dashboards.Rmd' describes the use of these processed datasets to produce two kinds of product:

• species-level summaries, averaging the complete set of seabed habitat / sediment properties over all known occurrences of the species (weighted by abundance). A complementary summary over all sample events is also produced;
• individual species plots, showing how a species' occurrences are distributed across both continuous (median sediment particle diameter, and percentage of mud, sand, and gravel) and categorical (Seabed Habitat Map Substrate, EUNIS classification, and MSFD classification) habitat variables. The distribution of these variables across all sampling events is also shown for comparison.

These workbooks are all rendered into both html and pdf reports in the docs folder.

Scripts

The three functions used to generate the products are included as separate scripts which are sourced in 'benthic species habitat dashboards.Rmd':

1. get_species_habitats.R gets an abundance-weighted species-level summary, averaging the complete set of seabed habitat / sediment properties over all known occurrences of the species (weighted by abundance);
2. get_event_habitats.R does the same as (1) but for all sampling events;
3. plot_species_habitats.R produces the composite plot for a given species, and either prints it, saves it to file (as an A4 pdf), or both.

Products

This contains the derived data products which are fully described in the Data Product section below.

Data Series

The data series used in these products are not easily obtained via WFS. The Biotic species traits data are provided in csv format within the raw data folder. The EMODnet benthic numerical abundance data product and the EMODnet broadscale seabed habitat maps are both very large downloads (>400MB zipped csv file and >500MB zipped ESRI shape file). Future work may add WFS download options, but in the current product these datasets are downloaded directly by supplying the dataset url to download.file function in R. The other dataset used (the sediment properties dataset) is also directly downloaded as a csv file in this way from its doi. The specific code to obtain these datasets is provided within the benthic data habitat matching R markdown document (in the Analysis folder), rendered as html and pdf reports in the Docs folder.

Data Product

The data product consists of two datasets, provided as csv files, and one synthetic figure per species:

1. benthic_species_habitat_summaries.csv for each of 3287 taxa a single row with some basic summary data, all habitat information from BIOTIC, and summaries of all the derived habitat variables - means (weighted by abundance) for continuous variables, frequencies (weighted by abundance) for categorical variables. In all there are 147 variables per species. These are described in the associated metadata file, benthic_species_habitat_summaries_metadata.csv.
2. benthic_sampling_event_habitat_summaries.csv contains similar information but averaged over all individual sampling events (1 row x 114 variables). Variables are named according to the same scheme described in benthic_species_habitat_summaries_metadata.csv.
3. The subfolder species_hab_plots contains a composite plot illustrating habitat affinities for selected species. Code for generating these plots for all species occurring in more than 20 sampling events is provided in the benthic species habitat dashboards document, and all 1844 of these plots are provided as part of the full published product. Each plot has the filename structure Aphia####_habitat_plot.pdf where #### is the Aphia ID of that taxon. Each plot is saved as an A4 pdf. The title of the plot gives the taxon scientific name and Aphia ID, the subtitle presents the number of sampling events in which the species occurs which have been matched to the different habitat variables. Habitat preference information from BIOTIC is also presented, if available.

More information:

References

• Beauchard, O., Herman, P. M.J., Fernández-Bejarano, S. (2021). Data product numerical abundance of benthic macroinvertebrates in North Sea and Baltic Sea. Integrated data products created under the European Marine Observation Data Network (EMODnet) Biology project (EASME/EMFF/2017/1.3.1.2/02/SI2.789013), funded by the by the European Union under Regulation (EU) No 508/2014 of the European Parliament and of the Council of 15 May 2014 on the European Maritime and Fisheries Fund. Available from: https://www.emodnet-biology.eu/data-catalog?dasid=6648
• MarLIN, 2006. BIOTIC - Biological Traits Information Catalogue. Marine Life Information Network. Plymouth: Marine Biological Association of the United Kingdom. Available from www.marlin.ac.uk/biotic
• Vasquez Mickael, Manca Eleonora, Inghilesi Roberto, Martin Simon, Agnesi Sabrina, Al Hamdani Zyad, Annunziatellis Aldo, Bekkby Trine, Pesch Roland, Askew Natalie, Bentes Luis, Castle Lewis, Doncheva Valentina, Drakopoulou Vivi, Gonçalves Jorge, Laamanen Leena, Lillis Helen, Loukaidi Valia, McGrath Fergal, Mo Giulia, Monteiro Pedro, Muresan Mihaela, O'Keeffe Eimear, Populus Jacques, Pinder Jordan, Ridgeway Amy, Sakellariou Dimitris, Simboura Mika, Teaca Adrian, Tempera Fernando, Todorova Valentina, Tunesi Leonardo, Virtanen Elina (2020). EUSeaMap 2019, A European broad-scale seabed habitat map, technical report. EASME/EMFF/2018/1.3.1.8/Lot2/SI2.810241– EMODnet Thematic Lot n° 2 – Seabed Habitats. https://doi.org/10.13155/74782
• Wilson et al. (2018) A synthetic map of the north-west European Shelf sedimentary environment for applications in marine science Earth Syst. Sci. Data, 10, 109–130 https://doi.org/10.5194/essd-10-109-2018
• WoRMS Editorial Board (2021). World Register of Marine Species. Available from http://www.marinespecies.org at VLIZ. Accessed 2021-04-01. doi:10.14284/170

Code and methodology

Explained above.

Citation and download link

Please cite this product as: Tom J. Webb (2021) Benthic occurrences, habitat maps, and species traits. Integrated data products created under the European Marine Observation Data Network (EMODnet) Biology project (EASME/EMFF/2017/1.3.1.2/02/SI2.789013), funded by the by the European Union under Regulation (EU) No 508/2014 of the European Parliament and of the Council of 15 May 2014 on the European Maritime and Fisheries Fund

Available to download in:

https://www.emodnet-biology.eu/data-catalog?module=dataset&dasid=6620

Authors

Tom J. Webb