Disturbance was computed by combining:

- Seabed shear stress (in N.m-2) based on Mediterranean modelled seabed shear stress (90th percentile)

- Sediment average grain size based on Modelled seabed sediment average grain size in the Mediterranean

Scope for Growth factors involved are:

- Surface Chlorophyll a (Chl.a ; mg.m-3)

Computed as maximum concentration of surface chlorophyll obtained from monthly satellite observations from 1998 to 2014 (MyOcean (http://marine.copernicus.eu/): "OCEANCOLOUR_MED_CHL_L4_REP_OBSERVATIONS_009_078").

- Bathymetry

EMODnet Bathymetry Consortium (2018): EMODnet Digital Bathymetry (DTM). http://doi.org/10.12770/18ff0d48-b203-4a65-94a9-5fd8b0ec35f6

- Stratification

Here, stratification was considered as the average absolute difference between surface and 30 (± 5) m depth density over 20 years. Salinity and temperature data used to compute density cover the 1994-2014 period and were from monthly model predictions (MyOcean (http://marine.copernicus.eu/): "MEDSEA_REANA-LYSIS_PHYS_006_004"). High Pressure International Equation of State of Seawater (1980) was used to compute Sea water Density. The description of the computation of density of sea water at high pressure from temperature and salinity is available at https://www.niot.res.in/COAT/coat_pdf/CHAP%20III%20-%20Equation%20of%20State.pdf.

- Bottom temperatures from 1994 à 2014 from monthly model predictions (MyOcean (http://marine.copernicus.eu/): "MEDSEA_REANA-LYSIS_PHYS_006_004") were used to compute average bottom temperature (BT), inter- and intra-annual bottom temperature standard deviations (Ti and Ta respectively). Ti, is the standard deviation between yearly averages and Ta, is the standard deviation between months averages.

- Average bottom disolved oxygen concentration (O ; mmol.m-3) from1999 to 2014, from monthly model predictions (MyOcean (http://marine.copernicus.eu/): "MEDSEA_REANALYSIS_BIO_006_008"). This was further transformed in % disolved oxygen using the equation of dissolved oxygen concentration in the water (after Weiss, 1970) found in JFE Advantech Co.,Ltd., Ocean and River Instruments Division, 2015 (http://ocean.jfe-advantech.co.jp/english/sensor/img/do.pdf)

R.F. Weiss, 1970. The solubility of nitrogen, oxygen and argon in water and seawater, Deep Sea Research and Oceanographic Abstracts, Volume 17, Issue 4, 1970, Pages 721-735, ISSN 0011-7471, https://doi.org/10.1016/0011-7471(70)90037-9.

The critical shear stress is the the minimal critical current needed to initiate sedimentary movement and was computed as a function of the log grain size (X, gain size in mm) and derived from the Hjulström diagram empirically approximated by the equation: Y= -0.0272 × X^4- 0.0905 ×X^3+ 0.2411 × X^2+ 0.4691×X+1.8761 and critical shear stress = (10^y)/100

Seabed shear stress (in N.m-2) was used to describe friction intensity.

All layers were limited to the same extent and rasterised at the same resolution (by bilinear interpolation).

The disturbance was then defined as

Dist = log(friction intensity/ critical shear stress)

followed by centering, standardisation and rescaling between 0-1.

Before compiling the different environmental layers to define the scope for growth, data were recoded between 0 and 1. Since this reclassification is sensible to the geographical extent of the studied area, all layers were limited to the same extent and rasterised at the same resolution (by bilinear interpolation).

The equation to calculate scope for growth is defined as:

SfG= (Fa+Temp+O)/3

with Fa, food availability, Temp, temperature impact, O, %dissolved oxygen. SfG was then centered, standardised and rescaled between 0-1

Food availability was computed as:

Fa = F - stratification

F=log(Chl.a/bathymetry), followed by centering, standardisation and rescaling between 0-1 and stratification was also centered, standardised and rescaled between 0-1 before computing Fa. Once computed Fa was also centered, standardised and rescaled between 0-1.

Temperature impact was computed as

Temp= BT-Ta-Ti, followed by centering, standardisation and rescaling between 0-1

%dissolved oxygen was expressed as ratio instead of %.

The process-driven sensitivity (PDS) can be seen as a risk map that combines the two previous axes, according the formula :

risk =sqrt[(1-SfG)²+(1-Dist)²]

followed by centering, standardisation and rescaling between 0-1