Figure 1.
Map of study area off southwestern Iceland (A) with sample locations indicated by open boxes.
Inset figures represent the proportion of cod exhibiting a particular gonadal maturity stage (1: immature; 2: maturing; 2.2: final stages of maturation; 3: ripe and running; 3.2: near spent; 4: spent/recovering; 5: omitted spawning) sampled from these sites during 17–20 April 2009 as part of the annual spring groundfish survey conducted by Hafrannsóknastofnunin. Scale bar represents 10 km. Bathymetric maps of Knarrarós (B) and Lofstadarhraun (C). Contour lines indicate 2-m isobaths. Scale bars in panels (B) and (C) represent 500 m.
Figure 2.
Example images collected from contemporaneously deployed ROV (top row), 120-kHz split beam echosounder (middle row), and DIDSON (bottom row) over boulder/lava field (a), transitional (b), and sand (c) designated habitats off southwest Iceland, 15–16 April 2009.
Note redfish Sebastes sp. identified in ROV video data collected from boulder/lava field habitat, Atlantic cod Gadus morhua identified in both ROV and DIDSON data, and haddock Melanogrammus aeglefinus identified in the ROV data with faint targets detected near the substrate in the DIDSON and 120-kHz split beam echosounder data.
Figure 3.
The dual-frequency identification sonar (DIDSON) views a horizontal plane of space, every point seen is described with a polar coordinate, (r, θ) within this (A).
The direction of view (B) is expressed as three angles: yaw (compass angle with 0 degrees as north), pitch (or tilt, the angle up or down relative to horizontal, with 0 degrees horizontal and −90 degrees looking vertically down), and roll (displacement of the observed horizon from horizontal). First the polar coordinates (r, θ) are converted to (x, y) in meters relative to the DIDSON. (x, y) is then rotated in three dimensions using the rotation transformation matrices corresponding to roll, pitch and yaw in turn to get a location (depth relative to DIDSON and displacement north and east).
Figure 4.
Least-square means estimates of acoustic fish density (fish nmi−2; upper panel) and mean acoustic target strength; lower panel) of fish at boulder/lava field (reef), sand, and transitional habitat surveyed off southwestern Iceland 15–16 April 2009 using a split beam echosounder with a 120 KHz 6° split-beam digital transducer operating at 0.4 ms.
Nearshore stations (white) and offshore (filled) stations are illustrated. Error bars represent upper and lower 90th percentile confidence intervals.
Figure 5.
Images of Atlantic cod captured by a dual-frequency identification sonar (DIDSON) deployed approximately 1.0 km off the coast of Iceland at Knarrarós on 15 April 2009.
In panels a–d, the DIDSON is positioned approximately 10 m off the bottom and is oriented at a 30° angle off perpendicular to the substrate. The Atlantic cod shown in these panels (a–d) are swimming up into the water column. The DIDSON is the same distance above the substrate in panels e–h, but oriented at approximately a 60° angle to it. These cod seem to have reached the depth that the fish in panels a–d were ascending to and are swimming parallel to the substrate. Arrows indicate two individuals that may be coupling and engaged in a spawning event. The full video file from which these images were captured can be found in the supplementary information (Video S1).
Figure 6.
Plot of height above the seabed of individual Atlantic cod in a presumptive spawning column over time as determined using a dual-frequency identification sonar (DIDSON) deployed off southwestern Iceland 16 April 2009.
Different colors represent individual fish. Figure describes the positions of fish seen in Video S1.
Figure 7.
Echogram of cod spawning aggregation in the transitional habitat between boulder/lava field (left) and sand (right) on Lofstadarhraun off southwestern Iceland 16 April 2009.