Fig 1.
An echogram collected during the North Sea trawl survey.
Acoustic data collected at 4 frequencies, 18, 38, 120 & 200 kHz, showing the seabed (region of strong scattering at ~140 m depth), fish schools (discrete regions of strong scattering across the four frequencies at ~100 m depth), and individual fish (areas of more diffuse scattering close to the seabed). These data were collected in proximity (<2 km) to one of the oil platforms surveyed.
Fig 2.
Maps of fish distributions and environmental data.
Maps of the North Sea showing: a) oil and gas platform areal densities, and the locations of platforms surveyed to within a 1 km range (black crosses); b) oil and gas pipeline areal densities; c) Aqua-MODIS satellited derived Sea Surface Temperature (°C) from August 2012; d) fish school density where the area of each circle is proportional to the Nautical Area Scattering Coefficient (m2nmi-2); e) density of non-schooling fish where the area of each circle is proportional to numbers per m2; and f) acoustically derived seabed habitat class (labelled 1–8) overlaid on the ship’s cruise track.
Table 1.
Details of platforms the survey vessel approached to within 1km.
Fig 3.
Boxplots of densities of individual fish (i.e. non-schooling) vs distance from platform.
Each subplot shows data closest to each of the 16 platforms surveyed at ranges <1 km. Data are binned to 500 m intervals; the x-axis values are the upper limit of bins (i.e. the box plotted at 1 km represents the data in 500–1000 m bin). Box colour represents the modal seabed class of the data for that box (see legend). Black lines are the lines of best fit of linear models, shown only where the relationship was significant (i.e. p<0.05). Plots are ordered by the model fit (R2 value). Boxes to the right of each subplot (marked * on x-axis) represent the equivalent ‘baseline’ fish density data; these are >25 km from any platform, from water depths within 10 m of the mean for each subplot, with the proportional distribution across seabed classes, and within 1° latitude of the mean of the data. The result of Welch tests for differences between the data in each box and the relevant ‘baseline’ are shown above the x-axis (with + and—signifying a significantly higher or lower mean than the baseline, respectively). Details of each platform (including water depth and platform type) are given in Table 1.
Fig 4.
Modelled relationships between fish density and explanatory variables.
Plots showing the shape of fitted explanatory terms found to be significant (at α = 0.01) during the generalised additive modelling process. Boxes a-c are terms from the model of fish school presence/absence, showing the relationship with a) platform density, b) depth and c) sea surface temperature (SST), and d-f are terms from the model of non-schooling fish density, showing the relationship with d) platform density, e) depth and f) bottom water temperature.