Last night there was an issue with the trawl net; when it was removed from the water it was significantly torn. This meant a new net, already calibrated, had to be fitted so we could carry on as usual.
Where we've been on the cruise so far, with key at bottom
Today saw us arrive in the Shetland Islands to refuel the boat and swap our sea legs for our land legs to look around the capital, Lerwick, for a few hours. Everyone had been greatly anticipating dropping in on the archipelago, and with the sun out in the morning, the journey into the port was a thrilling one.
Over parts of the survey area, the Johan Hjort has made its way along several transects scanning the depths for the fish, stopping intermittently when the echogram showed larger, redder patches (proportional to a larger fish) that the cruise leader considers of interest. Then a trawl takes place. 'Blind trawls' are also taken where, although there's little of interest being picked up on-screen, the scientists need to be sure nothing is being missed.
But the real-time tracking is only part of the picture. Acoustic readings for the nautical miles of the transect already cruised by the vessel need to be performed. This process – scrutinizing the acoustic data – is carried out twice a day on board by cruise leader Jennifer Devine and IMR instrument technician Jan Erik Nygård using the Large Scale Survey System (LSSS) software co-developed by IMR. Consistency is key here as experts can have different analytical techniques, and a switch of personnel could be reflected in the end data.
On the computer screen, the backscatter (energy pinged back towards and received by the vessel's echosounder) over the travelled path is displayed as coloured specks and patches. These are the reflections of the sonar waves off the part of the fish of the fish that the instrument detects: their swim bladders, which contain air and are therefore of a different density to the area around. Some blobs are shaped like tiny bananas because of the shapes of their swim bladders, the body parts that the sonar waves bounce off (as well as other reflective objects like the seabed, gas-filled bubbles and gas pockets in plankton and jellyfish; weak sound is also reflected off bones). To the untrained eye a lot of it looks like fuzz, but to the experts it is raw data ready to be picked apart in order to tell one sea species from another.
Sound waves from an echosounder bouncing back off an orange roughy, a species which doesn't reflect as strongly due to its swim bladder being full of wax esters rather than gas. The red and blue colours indicate a denser signal whilst blue and yellow a more dispersed one. The arrow here shows the bump of the swim bladder; waves also reflect of the fish's otoliths. Image: Macaulay, G. J., Hart, A. C., Grimes, P. J., Coombs, R., Barr, R., and Dunford, A. J. 2002. Estimation of the target strength of oreo and associated species. Final Research Report for Ministry of Fisheries Research Project OEO2000/01A Objective 1.
Because the screen mirrors how the sea is layered, a stratification that scientists call the water column, it is helpful to understand how this works in practice.
The echogram sliced into horizontal layers for each water column zone, the very bottom layer is a zoomed in version of the seafloor layer above it. The left scale is water depth, the right the backscatter – amount of pinged back energy – threshold.
The transects are broken down into clearer-to-see segments of five nautical miles, to historically go back over the cruised area to determine what organisms were in the echosounder's radius around the boat – and their composition.
The boat's path, along which the acoustic data is scrutinized
So how to single out saithe from the acoustics? It's a stealthy creature, hard to pick up using underwater sonar, so it's not always an easy task. Telling it apart from other big fish is also tough, a fact not helped by the acoustic 'dead zone' adjacent to the sea floor where demersal fish may be hidden by the echo of the sonar waves off the bottom.
We know saithe is benthopelagic – lives close to the bottom (the benthos) but also venturing into higher waters (pelagic zone). So a range of around 100-200 metres, although it has also been caught in surface tows. To start with the three regions are set on the screen: lower (typically the 10 metres closest to the sea floor), middle (usually pelagic fish and plankton) and upper (usually herring and densest plankton congregations). The backscatter threshold is increased, which filters out the plankton from each region in turn.
In the midwater (pelagic) zone, once the plankton are out, bigger fish that are left can be selected with the mouse and merged with the demersal layer. The sonar has also picked up lots of young haddock and whiting (known as zero group due to their age) in this zone this year, and these are removed in the scrutinizing process as they're not target species.
Here's the echogram after plankton has been removed from each of the layers (compare with screenshot above). The peaks in the second red line down indicate where large fish in the middle area have been identified and ringed in order to include them in scrutinizing the bottom layer.
Further work now need to be done in the bottom (demersal) zone, the region of interest. Sometimes there are dense red patches representing Norway pout. At night, as the fish rise from the sea floor and disperse. If this happens, they too can be disregarded along with the pelagic zone. If they're on the seabed, they're left in the layer.
Also, the fish sampled in the catch at a trawl site is vital in determining species composition of the backscatter. Data collected from the fish caught be the trawl can be displayed alongside LSSS, where number of each species caught, total weight, length frequency information, as well as proportion of the backscatter attributed to the main species can be displayed. Looking at this data, as well as the lengths, can help when scrutinizing.
The fish by species, total weight and number of individuals, as caught at each of the acoustic trawl stations.
Further clues to the species composition can be given using the frequency response – the amount of backscatter at each of the four acoustic frequencies used on the vessel. When a dense school in the bottom layer could be either herring or Norway pout, for instance, a ring can be drawn around the patch to find out which, as the species have very different responses. This can also help for identifying saithe.
Using the data from the trawls carried out within a certain radius coupled with the acoustic information, we then can work out the proportions of backscatter that should 'belong' with each species.