Food-web data
We studied the trophic network corresponding to the upper part of the Celtic Sea trophic network (TL>=2), defined at genus or species resolution and restricted to interactions between most abundant and common taxa observed during the scientific survey EVHOE (see Mérillet et al., 2020). Trophic links between taxa were taken from the literature review conducted in Hernvann et al. (2020), keeping the information at the taxon’s level. We collected isotopic measurements for 69 taxa of the most commonly found in bottom trawls in the Celtic Sea (EATME project). We applied a lipid correction for the taxa with a C/N >3.5, following Sweeting, Polunin, & Jennings (2006). We then applied a baseline correction based on isotopic signatures ofAequipecten opercularis , using bottom temperature for δ13C (Barnes et al. 2009) and using bottom temperature and salinity correction for δ15N (Jennings & van der Molen 2015). The performance of the correction was then checked by looking at the correlation between real and corrected values for other bivalves (i.e. Pecten maximus , Aequipecten opercularis , Pseudamussium peslutraeand Pseudamussium clavatum ). Trophic levels TL of taxa i were computed from Post (2002) equation, as follows:
\(\text{TL}_{i}=\ \frac{\delta{15N}_{i}-\ \delta{15N}_{\text{base}}}{3.4}+\text{TL}_{\text{base}}\)
with δ15Ni the corrected δ15N value for the taxa i and δ15Nbase the mean of the δ15N base values of bivalves (TLbase = 2).
We computed several metrics informing on the structural properties of the network. Connectance (L/S2), computed from the number of links (L) and the number of species (S), gives information on how densely connected is trophic network. Our network of 69 species and 559 links has average values of connectance (0.12), with connectance ranging from 0 to 1, with 0 indicating the least possible number of interactions (Delmas et al. 2019). For a complete description of these metrics see Delmas et al. (2019) and Kortsch et al. (2018).