Isabel Key, CC BY-ND " src="https://s.yimg.com/ny/api/res/1.2/StTG05MvauDElpLtNdpavQ-/YXBwaWQ9aGlnaGxhbmRlcjt3PTk2MDtoPTcyMA-/https://media.zenfs.com/en/the_conversation_464/5e8ad10ecacd497f7e486f9daf023b6c" data-src="https://s.yimg.com/ny/api/res/1.2/StTG05MvauDElpLtNdpavQ-/YXBwaWQ9aGlnaGxhbmRlcjt3PTk2MDtoPTcyMA-/https://media.zenfs.com/en/the_conversation_464/5e8ad10ecacd497f7e486f9daf023b6c"/>
Scotland's rugged west coast looks stunning in the sunshine. The turquoise sea is calm, but even in July it's a chilly 12°C. Armed with my recording kit, snorkelling gear and my thick wetsuit, I listened to the marine life that lives in three different Scottish seagrass meadows.
For my PhD in marine ecology I am studying the biodiversity of Scottish seagrass meadows which are now in full bloom in midsummer. Unlike seaweed, this marine plant has flowers, seeds, pollen and underground roots.
Seagrass meadows are abuzz with activity. Sea snails scrape across rocks as they eat algae, young fish feed on tiny zooplankton, crabs fight to defend their territory, and birds, seals, and otters hunt for food.
There's a cacophony of sounds from all this activity, and I'm researching how seagrass soundscapes - the collections of sounds heard in an environment - differ depending on the animals that live there. Hearing a greater variety of sounds can mean that there are more animals living in the seagrass, and can potentially indicate a healthier, more biodiverse seagrass meadow.
Seagrass meadows have declined dramatically due to sediment and nutrient runoff from agriculture, coastal development, destructive fishing practices and disease. The UK has lost over 40% of its seagrass cover, possibly up to 90% compared to pre-industrial levels. Globally, 29% of seagrass has disappeared since the 18th century and the rate of decline has accelerated, with around 7% lost per year since the 1990s.
Seagrass is an important breeding ground for fish, it improves water quality and acts as a carbon sink. Its decline therefore affects the marine animals that live in the habitat, animals further up the food chain and the health of the ocean in general.
Recording soundscapes in seagrass is useful because it allows researchers like me to detect creatures that we can't necessarily see, because they're camouflaged or hiding, or maybe they're nocturnal. It also causes minimal disruption compared to other monitoring methods, and it could become cheap and efficient. In the future, you might just be able to put a recorder down, pick it up, run some algorithms, and get some information about the animals that are there.
Listen to Isabel Key explain the sounds she collected in an interview on The Conversation Weekly podcast.At each meadow I visit, I set up palm-sized underwater microphones on stands and leave them on the seabed for a week. Every day I snorkel down to set up a video camera next to the microphone so I can match the sound to the video. That way I can figure out which sound is being made by which animal.
Back at the office, I analyzed my audio recordings using "acoustic indices," which measure the complexity of the soundscape. This includes animal sounds, but also waves, boat noises, and creaking mooring chains.
Next, I assess the richness of sound by listening to one-minute clips. When I look at the spectrogram - a visual representation of these sounds - I can count how many different types of animal sounds are present. This takes time, but gives great insight.
So far I've identified 14 different types of sounds that I suspect are coming from fish and crabs living in the seagrass, plus dolphin whistles and clicks that I can hear from further away as they swim by. I can look at the exact frequencies (or pitch) that sounds hit and the patterns they make, and then more accurately attribute that sound to an animal or human activity.
I found some evidence for a characteristic seagrass soundscape with certain sounds occurring more frequently in seagrass than in sandy habitats. Fish make low grunting, burping or purring sounds. Crabs make higher pitched, metallic scraping sounds.
I often hear a popping sound that becomes more pronounced as the day progresses. As the seagrass photosynthesizes, especially in the middle of the day when the sun is hot and bright, the plant produces oxygen bubbles that collect on the surface of the seagrass leaves and pop as they enter the water.
It's hard to decipher which animal is making which sound, especially since our oceans are such noisy places. Acoustic pollution can be a serious problem for marine animals that rely on sound to survive, find a mate, navigate, communicate with each other, or hunt for food.
Interestingly, seagrass may act as a buffer against underwater noise pollution. As a 3D structure, seagrass acts as a physical barrier against wave energy-one reason it plays a crucial role in protecting coastal areas from erosion. It can also absorb sound waves and even protect fish from dolphins that use echolocation to navigate to their prey. Dolphin clicks don't penetrate seagrass very well, so fish may be safer in this seagrass sound shelter than they are in the open sea.
In two meadows I found more fish and crab sounds in the seagrass than in the sandy sites I compared them to, just as expected. But in one location I heard more sounds over the sand than in the seagrass, despite the lower wildlife. So biodiversity levels are not necessarily directly reflected by the soundscape.
This may be partly due to differences in how sound travels in different habitats. Sound travels more easily through sand than through seagrass. This phenomenon can lead to misleading results where it is harder to hear fish in denser seagrass because the seagrass itself absorbs the sound, even if there are more fish living in it.
Researchers should be careful when interpreting soundscape data and consider how habitat structure affects how sounds are heard. Acoustic monitoring may therefore be more useful for studying changes in animal life over time in a single location than for comparing across different areas.
Pasture monitoring
The hope is that this kind of work can be used to train machine learning algorithms and eventually build an easy-to-use tool for monitoring biodiversity in seagrass and other marine habitats. For that, a large library of sounds is needed. This already exists for dolphins and other marine mammals, but is not yet well established for sounds from fish, crabs, and other invertebrates such as shrimp.
Capturing all the different sounds each species can make usually starts with recordings in an aquarium. Then automated detection can try to match those with sounds that researchers like me record in the field. This should allow scientists to identify early indicators of meadow decline or measure the success of seagrass restoration projects.
Perhaps one day marine scientists around the world will install sound recorders that can transmit audio samples from coastal oceans to a central online database where soundscapes can be automatically analyzed to assess ocean health. This could give us near-real-time data on animal populations and movements, helping to inform marine conservation measures and sustainable fishing practices. That's an exciting prospect.
This article is republished from The Conversation under a Creative Commons license. Read the original article.
