Over the holidays Marine Scotland announced they had found a new deep-sea sponge reef on the extinct submarine volcano known as Rosemary Bank in the Rockall Trough, North East Atlantic; and from the imagery provided by Marine Scotland the sponge reef looks to be of a type dominated by a species known as Pheronema carperteri aka the bird’s nest sponge.

Rosemary Bank Sponge Reef. Image owned by Marine Scotland

When I first heard this news and saw the images I was so excited I nearly got my laptop out straight away, but as it was the Christmas holidays I thought better of it. Why was I so exited? Well apart from the fact the deep-sea sponge reefs are extremely cool, very important to the deep-sea ecosystem, and fascinating to boot, this new discovery potentially provided me with data to test some of the predictive maps my research group have been creating over the past couple of years. What is a predictive map? I’ll get to that.

You see the deep sea is big, really big, vast if you like, and one of the biggest problems we have with trying to look after it properly (given that we seem to be okay with exploiting it for food and raw materials) is in understanding where important habitats like deep-sea sponge aggregations are likely to be found so that we can avoid damaging them. Sometimes we stumble across them in the course of our research as Marine Scotland did with this reef. Once we know they are there we can take action to protect them from potentially damaging activities. However, this means that adequate protection of deep-sea habitats moves at the pace of our exploration of this vast ecosystem. We have all heard the stock (yet true) phrases like “we know more about the moon than we do about the deep sea”, and “we have explored less than 1% of the deep sea”. If we continue to only concern ourselves with areas we have actually looked at, then deep-sea conservation and sustainable management is not going to much progress any time soon.

What would be really useful is if we had maps of where all the different types of deep-sea habitats are found, then we could make some strategic decisions on where we will allow certain human activities to go on and where we will not. The question is how can we create such maps without looking everywhere in the deep sea? Well, all habitats and species only occur under the “right” environmental conditions for them. You won’t find a mangrove in the Arctic, and you don’t find polar bears in the desert (I am being extreme but you get the idea). If we know the environmental conditions a habitat occurs under, we can predict where else we should find that habitat based on whether another area has the “right” conditions. This is called predictive habitat mapping or modelling. Essentially it uses some reasonably complicated maths to work out the probability of you finding your habitat of interest in an area based on how “right” the environmental conditions are for it in that area.

So over the last 2 years my team have been producing predictive models (maps) of where we should find deep-sea sponge reefs in the North East Atlantic. We have been looking at 2 types of sponge reefs, a type called ‘Ostur’, also known as cheese bottoms, and a type dominated by the bird’s nest sponge Pheronema carpenteri – the type of reef recently found by Marine Scotland.

Clockwise from top left, Ostur sponge habitat, Pheronema carpenteri aggregation, close up of Pheronema, close up of Ostur. Images Crown copyright 2006

When I heard that Marine Scotland had found a sponge reef dominated by Pheronema carpenteri on Rosemary Bank my first thought was did our models predict it should be there or not? I hastily plotted up data from two of our models that were built using slight different accuracy data. One was built using data at a resolution of 750m, the other at a resolution of 200m. I won’t go into this aspect of the modelling now but using data of different resolutions gives different maps so we wanted to check both. Yes our models did indeed both predict that we should find this type of sponge reef on Rosemary Bank, but was the area Marine Scotland found the reef where we predicted it should be. Well we don’t have the exact position data yet but aligning their survey image with our maps strongly suggests they did find the reef exactly where both our models predicted it should be.

Clockwise from the top left, an image of where on Rosemary Bank Marine Scotland carried out their survey, the area of Rosemary Bank our 750m model predicted the deep-sea sponge reef should be, the area of Rosemary Bank our 200m model predicted sponge reef should be, all three maps overlaid to show the area Marine Scotland surveyed aligns well with where both models predicted sponge reefs would be.

Is this exciting? Absolutely! If we can make accurate maps of where different deep-sea habitats might be found using predictive modelling we can really start to plan how we use the deep-sea in a more informed way.

We won’t know how good our maps are until we have had a chance to test them properly with this new data from Marine Scotland. Plus this summer our group will spend two and a half weeks at sea gathering new data to test these and other modelled maps that we have created in order to begin to fully understand whether this approach offers us a useful way forward. But for now it certainly looks like our maps predicted there would be a deep-sea sponge reef on Rosemary Bank and there is!