Tuesday, 24 July 2012

Day 19 - The Abyssal Plain. Not so Plain after all

This is the main event! We mentioned before that Autosub6000 was being prepared for duty, and after a furious couple of weeks of work at Porcupine Abyssal Plain, the sub has successfully completed seven missions, collecting bathymetry data and over half a million photos of the seafloor.

Photo strip showing a dumbo octopus and a sea cucumber
These surveys are designed to gain a broader perspective of how communities might differ between abyssal hills and flat areas, which is similar to understanding how communities of animals might change as one moves from a valley to the summit of a mountain. On land it is easy to confirm on the ground what we can see from satellite photographs, but this is far harder in the deep-sea because we can’t see through the water using light. Instead, we have to use sound-based mapping techniques such as sonar to map large areas of the seabed and confirm them with remotely taken samples like those from the megacorer or photographs. The benefit of using Autosub for collecting photos of the seafloor is that it can cover vast stretches of the seabed quickly while collecting a huge number of photos, along with other data about the environment. The photo transects we’ve collected during this cruise span an area about the size of a medium-sized city, enabling us to view a greater area of the abyssal landscape than has been generally possible for deep sea areas.

Autosub has been deployed with two cameras. The downward facing camera takes photos directly below Autosub, and these photos are used to assess the number and types of invertebrates that live on the seafloor in different areas of the plain. Sea cucumbers are the dominant type of large invertebrates that we see in the abyss, and there are at least ten different varieties in our photos. Autosub takes one photo every 0.8 seconds, so the photos overlap. This overlap means that photos can be stitched together into long strips, giving a continuous picture of the seafloor. In just a few days, we have taken more than 300,000 photos with the downward camera, which will take months to analyse!

The forward facing camera is being used to look at the numbers and distribution of abyssal fish. Fish decline in number rapidly with depth, with only a few species able to survive on the limited food supply in the abyss. The rarity and mobility of fish means that surveys must cover large distances to determine their distributions over the seabed, which is why Autosub is such a valuable tool. Fish species in the abyss look quite different from fish that you see at the fishmongers; they are dominated by eels and rattails, and are typically scavengers that can cover long distances using relatively little energy to try and find food. We need to use a forward-facing camera to monitor fish because they are mobile, so can be startled and swim away before being seen by the downward facing camera. So far, we have collected around 250,000 photos, which means lots of fish to look for!

A steep volcanic rock face on the abyssal hill
In addition to Autosub, we have used SHRIMP (Sea floor High Resolution Imaging Platform) a towed camera system, to photograph one area of cliffs that was too steep for Autosub. The SHRIMP system had forward and downward video cameras and a downward facing still camera. Using this, we were able to see some rare rocky outcrops in the abyssal plain! This was really exciting, since this environment is largely flat and covered in muddy sediment.

As we come to the end of the highly successful Autosub missions at the Porcupine Abyssal Plain and anticipate exciting new finds, we find that we are positively swimming in photos!