(from final report):

Experimental Cage
A cage (4 m wide x 5 m long x 4.9 m deep; approximately 100 m3) with 5-cm (2-inch) stretch nylon mesh, which has served as a harvest/transfer cage for the OOA project, was used for a series of acoustic measurement experiments (Figure 2). The cage consisted of a high-density polyethylene pipe of 10.2 to 25.4-cm (4 to 10-inch) diameter for the top frame, rail and flotation. A 0.9 m (3-foot) wide wood boardwalk surrounded the cage and supported multiple people and docking vessels. The net was covered with a top net and lowered by rope to any depth. The bottom of the net was weighted by a metal rectangular frame. The floating net cage was towed from Portsmouth harbor and moored offshore at the OOA site with four mooring lines.

Video Surveillance

Two LED underwater cameras were mounted through the mesh net of the cage to provide an upwardlooking
and side-looking video record of the spatial distribution and behavior of the insonified fish during
the experiments. Each camera recorded to a four-channel digital video recorder (IDView).

Acoustic Sampling
A 300-kHz EM3002 multi-beam sonar system (Kongsberg Maritime) and 38- and 120-kHz Simrad EK60 split-beam echosounders were used to collect acoustic backscatter of the fish in the cage. The EM3002 single head has 160 beams each with 1.5° circular beamwidth that can collectively cover a 130° sector. The circular beamwidth of the 38 kHz and 120 kHz transducers are 12° and 7°, respectively. The multiand split-beam transducers were mounted on a rigid pole with the EM3002 single head in the center and the 38 kHz and 120 kHz transducers mounted on either side. The transducers were center aligned according to outside physical dimensions. The transducer mount was lowered approximately 1-2 m from a bridge across the center of a fish cage (Figure 2 and 3). A profiling sound velocimeter (Odom Digibar-Pro) was used to periodically measure sound velocity profiles and upload the sound velocity profiles to the EM3002 software (Figure 4). The sound velocity probe was attached near the single EM3002 head to collect real-time sound velocity measurement at the same depth of the head for proper beam forming. Instruments transmitted data via cable to vessels tied up alongside the cage. The 10-m vessel R/V Cocheco, both operated by the Center of Coastal Ocean Mapping and Joint Hydrographic Center, provided ship support for sonar operations (Figure 5).

Calibration
Sonar systems were calibrated by standard sphere calibration as described by Foote et al. (1987, 2005). A 38-mm tungsten carbide sphere, 23-mm copper sphere, and 60-mm sphere were used to collect single TS for calibration of the 300 kHz EM3002 sonar, 38 kHz E60 and 120 kHz EK60 echosounder, respectively. For each transducer, a sphere attached to monofilament line was lowered by a fishing rod from the sonar mount platform to a depth of 8-10 m and above the lowered cage. For the EM3002, the sphere was also lowered in other beams. The difference between the mean TS and the known TS for the sphere was adjusted by setting a gain offset.

Experiments
Before stocking the cage with live fish, acoustic measurements were obtained on the empty cage to determine whether there was acoustic transparency of the net or whether the cage would be well-defined to permit experiments to be conducted with the cage submerged or at the surface with the net cover open. After determining that the net formed discrete top and bottom echoes in the split-beam echosounders’ echograms and all four sides were distinct in the water column image in the EM3002 (Figure 5), all experiments were performed at depth with the cage completely sealed. Live adult cod (n=195) were stocked in the empty experimental cage at the surface at four stocking densities (approximately 0.25, 1, and 2 fish per m3) starting with the highest density. Fish were insonified in the cage from the surface by the three synchronized echosounders at 1 ping per second. Acoustic measurements on the largest cage population continued on the first day through 2 hours past sunset to provide a preliminary assessment of diel effects on the acoustic estimates of density. Depth of the bottom of the cage was also manipulated between depths of 6 and 17 m.