The Fish-edge project combines scientific diving, eDNA technology and underwater video surveys to explore coastal ecosystems at the edge of warming and invasion from non-native species

Coastal Mediterranean ecosystems are among the most threatened by warming waters and an influx of non-native species. More than 100 non-indigenous species originating from the Red Sea have now colonised the Mediterranean Sea through the Suez Canal, transforming ecosystems across the region.

Understanding how warming impacts both indigenous and non-indigenous fishes is important for better understanding the consequences of our changing seas.

Sébastien Villéger, research director at CNRS (EMBRC France) and the Marbec (MARine Biodiversity, Exploitation, and Conservation) lab at the University of Montpellier is an ecologist interested in temperate reef fishes. He and his four colleagues Victoria Delannoy, Nicolas Loiseau, Franck Ferraton, and Laure Velez) wanted to learn how warming and the arrival of non-indigenous species impact the waters off Crete.

Thanks to AquaServ’s transnational access programme, Villéger’s team was able to benefit from the facilities and expertise he needed at no cost.

The five French researchers collaborated with two Greek scientists (Thanos Dailianis and Grigorios Skouradakis) and accessed scientific diving support from EMBRC Greece (HCMR) which enabled them to explore this question using environmental DNA (eDNA) techniques, remote underwater paired cameras (RUPC), diver operated video (DOV) and sampling. Here, Villéger shares what they learned.

After the lockdown, our contacts at HCMR (EMBRC Greece) spotted more and more non-native species in the waters off Crete. In just a few years, new species were becoming abundant.

Working together with HCMR, our goal was to quantify the number of indigenous and non-indigenous fishes in five coastal Cretan habitats and discover whether warming impacts their biology, growth, stress levels, and grazing rates.

In particular, we were interested in non-native fishes that have been increasing in abundance in recent years (e.g. Parupeneus forsskali and Torquigener flavimaculosus) as well as a herbivorous species of rabbitfish (Siganus spp.) that overgrazes on algae. In some Cretan coastal habitats, rabbitfish are now responsible for more than half of all grazing activity. Their feeding pressure can strip rocky reefs of algae, creating barren seascapes often described as “underwater deforestation”.

We don’t know exactly. What’s sure is that the Suez Canal is an open door from the Red Sea. From there, it’s likely they started colonising the Levantine basin (Israel, Lebanon, Turkey) and may then spread to the Aegean Sea. Fish larvae can travel thousands of kilometres and might even be able to raft under natural or plastic material to colonise other spots.

Environmental DNA (eDNA) is the genetic material (such as mucus, faeces, skin, and gut tissues) shed by organisms in the water column. To collect these pieces of DNA, we used an underwater pump, carried by a scuba diver who swam along the reef two metres above the seafloor for 30 minutes. The pump filters seawater directly underwater, capturing traces of DNA on specialised filters. Those are processed when back in the lab, so the DNA can be amplified and then sequenced in order to identify the species present at each site.

A colleague at MARBEC lab has been building the most complete DNA reference barcode dataset for Mediterranean fishes and recently added 100 species from the Red Sea. This technique has helped us identify lots of species, proving that eDNA is useful for surveying all species present, even when they are scarce or shy. This is particularly important at the early stage of an invasion.

We used video to complement our eDNA work. The footage was taken at the same site, same time of year (early October), and in exactly the same spot as six years ago for comparison.

Two scuba divers recorded footage while swimming with an underwater camera for 50 metres: one to record fishes and the other filming the habitat along the transects. We filmed at five sites and in all except one which is close to a scuba diving centre (so the fish were used to the presence of humans), the fish swam away, making it hard to identify them on camera. This shows that diver operated video techniques have bias, so we also deployed remote video cameras on the seafloor.

The abundance of freshly arrived species was incredibly striking. The small pufferfish (Torquigener flavimaculosus) were abundant in four of the sites and we saw tiny juvenile Red Sea goat fish (Parupeneus forsskali) as well as large adults so we know they’re able to reproduce.

We sampled the fishes’ scales to measure stress and growth rates and compared them with those of the same species in French waters, where it’s much cooler on average. The analysis will be completed soon but we expect to see a faster growth when waters are warmer.

We’re also developing deep learning algorithms that we might use later for video analysis. This is an increasingly effective way of detecting and identifying fishes from footage, making data analysis much more time-efficient.

Our eDNA and videos can also provide baseline data: we can store them as a snapshot of what the seafloor was like when they were taken. Maybe in 10 or 20 years’ time, we can compare the historical record with the future situation and compare the two using AI – making our fieldwork have an even longer-lasting impact.

I heard about the Transnational Access AquaServ call through an email from a colleague. The application was easy through the online portal and we got the positive answer pretty quickly. It was really easy to organise the campaign.

Working with EMBRC Greece was excellent. We’d already collaborated with two colleagues six years ago so we were building on a fruitful partnership.

Our HCMR colleagues have the regional knowledge and scientific expertise to select the most suitable sampling sites and facilitate access. This is really important on such a big island which has several habitats colonised with different species. HCMR also provided scuba diving gear and support to carry out our field work.

I would recommend AquaServ TA funding for supporting collaboration.

Climate change and human activities are threatening ecosystems in many ways: one is the ability of some species to colonise new areas. Some of these invaders remain scarce and have little effect on the habitat while others become highly abundant, with a potentially catastrophic ecological impact.

For example, when rabbitfish overgraze on algae, it changes the habitat and food source for all the other herbivorous species and can have a ripple effect up the food chain. It’s important to understand which colonisers are able to successfully establish themselves and which might pose a threat in future so we can act quickly when interlopers arrive in the Med.