Sunday, July 24, 2016

Can you hear me now?

Figuring out the loudness at which an animal is vocalizing is a deceivingly difficult question to answer.  To investigate this question of loudness, we use an equation that, on the surface, is deceivingly simple – the sonar equation. 

SL = RL + TL

Source level (the actual loudness of the call) is equal to the received level (the loudness of the call when it gets to the hydrophone) plus the transmission loss (the amount of loudness that’s lost over the distance between the caller and the hydrophone).  How bad could it be?

Calculating received levels – this involves a few pieces of information.  First, you need to know when your animal is actually vocalizing.  I’ve browsed through dozens of hours of acoustic data from the 2015 season and have marked hundreds of calls from harbor seal males.  Lucky for me harbor seals call almost nonstop; at least finding the calls is easy enough.  I also need to know some information about my hydrophone, mainly the sensitivity.  Different hydrophones have different sensitivity levels, which influences the calls it can pick up.  A more sensitive hydrophone can pick up quieter calls than a less sensitive hydrophone.  Then, after I’ve input these hydrophone parameters, with the flick of my magic acoustic wand, a program called Raven will give me a received level.  Step one is complete.

Pretty picture from Strawberry Island to break from the scienceness of this blog post (Photo: L. Matthews)

Transmission loss is a bit more of a beast.  Back in the 1980’s, Malme and Miles came to Glacier Bay and investigated how sound travels underwater in this environment.  Turns out that sound attenuates at a rate of 15log(r), where r is the distance from the vocalizing animal to the hydrophone.  So if I want to quantify how much loudness is lost between my animal and my array, I need to know from where my animal is calling.  Enter acoustic localization. 

Sound is a pressure wave traveling through a medium.  In water, sound travels approximately 1500 m/s.  Depending on where an animal is located when it calls, the vocalization will travel through the water and arrive on the different hydrophones in our hydrophone array at different times.  The call will get to the closest hydrophone first, and the farthest hydrophone last.  You can see what I mean in the spectrogram below.  Each line is a different hydrophone, and this particular calling animal is closest to the second hydrophone.


I am lucky enough to be collaborating with a group at Cornell University called the Bioacoustics Research Program.  They’re the ones who make the software that I use to do the acoustic localization.  It’s been an up and down process to get this localization up and running, but after a few rounds of troubleshooting, it’s working with a fair amount of consistency.  It’s a slow and time-consuming part of calculating source levels, but so far I’ve managed to locate over 250 harbor seal vocalizations. 

Harbor seals called here.

Then there’s some coding involved to get the actual distances between the caller and the hydrophone, and a little excel spreadsheet organizing, but then it’s just plugging and chugging to get a number for a source level.  Sweet science victory. 

Hello sweet angel.  (Photo: L. Matthews)

I’ve always been intimidated by calculating source levels, and rightfully so.  There are a lot of different pieces that have to fall into place for everything to turn out correctly.  In my downtime on Strawberry Island, I’ve been slowly but surely chipping away at my source level analysis.  The island seems to be a good place to hunker down and focus.  Somehow it’s so much easier to be on my computer and not get distracted when the internet is just a distant memory….

Working hard on the beach.

Now that I’m getting baseline estimates of how loud harbor seals are vocalizing, I can start to answer another question.  Do harbor seals change their source levels when there are vessels nearby?  We see this phenomenon in many species, including humans.  When our environment is loud, we get louder so that other people can hear us.  I’m curious to see if harbor seals do the same thing.


Cruise ships are a regular occurrence during the summer (Photo: L. Matthews)


FUN SCIENCE FACT #36: Sound travels almost 5x faster underwater compared to in air.  From the hydrophones, we can hear a cruise ship coming almost 45 minutes before it gets to the island.



Friday, July 15, 2016

Mysterious little harbor seals...

So far, the harbor seal data collection has gone about as good as could be expected for any new protocol.  I had a few set backs early on with equipment, which have since been resolved, and there’s also been a lot of forward progress, which is great.  The part I’m struggling the most with, actually, is the part I thought wouldn’t be a problem at all – where are the seals that are vocalizing underwater. 

Harbor seal in the foreground, glacier in the background (Photo: L. Matthews)

I looked at the acoustic data from last year, and harbor seals are vocalizing all the time.  During the breeding season, male harbor seals set up underwater territories and defend them acoustically – makes sense that they would vocalize all the time.  I kayaked around our survey area the first few days we were here, and harbor seals are popping up everywhere – seems like there would be a lot of territories. 

In an ideal world, I drop a hydrophone from my kayak near where an individual is vocalizing, that individual vocalizes for a bit and then pops his head up out of the water to take a few breaths.  When that individual is at the surface, I take his photo and his position.  If I do this enough times, I can get a general idea of the size of size of this animal’s territory and compare that to the locations of the vocalizations underwater.   I can also associate the vocalizations I’ve recorded to specific animals and look at variation in vocalizations between individuals.

The view from my kayak office is kind of amazing (Photo: L. Matthews)

After some days of dropping the hydrophone and listening, I found a few places where the harbor seals roared loudly.  I sat in these spots and watched and waited for a harbor seal to pop up within sight range of the kayak, but nothing!  Multiple times now, we’ve made recordings of harbor seals, and we’re definitely close to where they’re vocalizing, but they never actually make an appearance at the surface.

I don’t know why it appears that my vocalizing animals never come to the surface.  I’ve got a little tinkering to do with the data that might clue us in as to where they are, but as of right now, your guess is as good as mine.


FUN SCIENCE FACT #45: Harbor seals may not be very agile on land, but in the water, they can swim up to 15mph to escape predators.  Most of the time though, they cruise around at a much more casual pace.

Harbor seal swimming gracefully in the cold waters of Glacier Bay (Photo: L. Matthews)



Friday, July 8, 2016

Always Listening, Sometimes Watching

Everyday the Strawberry Island field team wakes up and starts off the morning watching either whales or seals.  We take a break, watch some more whales and seals, take another break, and end our evening watching whales and seals again.  And throughout all of our data collection of marking what the animals are doing at the surface, our hydrophones are recording what the animals are saying underwater.  We do our best to try and understand what the whales and the seals do at the surface during all times of the day, but even at our best we can only watch some of the time.

Humpback whales are a common site near Strawberry Island (Photo: L. Matthews)

Acoustics is a great way to monitor animal behavior for a variety of reasons.  One of these reasons is that we can leave our hydrophones underwater to record animal vocalizations for months at a time.  No matter the time of day or type of weather, the hydrophones are archiving the acoustic behavior of all the whales and seals in and around the array.  And then, when we pick up our hydrophones at the end of the season, we can correlate the behavioral data we collected all summer to the acoustic data.  Our hydrophones are arranged in such a way that by using recordings from all four of them, we can determine the location of the vocalizing animal.  Acoustic localization plus behavioral data equals a lovely picture of what's happening in the survey area.

Spectrogram of a harbor seal roar recorded in Glacier Bay
Harbor seal (Phoca vitulina) in Glacier Bay (Photo: L. Matthews)

Understanding the correlation between what the animals are doing and what they are saying is an important link in the field of animal bioacoustics.  If we can relate these two datasets – the visual observations and the acoustic data – it will hopefully give us insights into what the animals are doing even when we’re not watching. 



FUN SCIENCE FACT #44: The clocks we use to synchronize our hydrophones for localization are among the most precise clocks in the world.  We're talking military-grade clocks (because that's a thing).  The only more accurate clock is the one on your cell phone. 


Sunday, July 3, 2016

Alaskan Adventure

Eighteen days ago I was dropped off on an island in Glacier Bay National Park with five other people, a dozen tarps, a fair amount of scientific equipment, thirteen 5-gallon water jugs, and 36 bear cans filled to the brim with food appropriate for a camp stove.  The goals: shore-based data collection on the behaviors of humpback whales and kayak-based data collection on the surface positions of harbor seals.

This fieldwork is part of my PhD research (the harbor seal part, at least), and it’s related to the hydrophone deployment I wrote about last year.  What I failed to write about more recently was that we deployed those same hydrophones in the same general area for a second year of data collection.  These hydrophones sit on the ocean floor all summer and listen to the sounds of harbor seals, humpback whales, and vessels.  The data from these hydrophones allows us to assess how noise from passing vessels affects the vocal behavior of two of Glacier Bay’s marine mammal species.  Last year there was a field team on this island collecting humpback whale behavioral data for the dissertation of my Glacier Bay counterpart, Michelle Fournet.  This year, I’ve tagged along with the humpback whale team to organize a harbor seal data collection initiative that will beef-up a chapter of my own dissertation. 

Harbor seal (Phoca vitulina) in Glacier Bay (Photo: L. Matthews)

The island is Strawberry Island.  It sits west of the Beardslee Island complex and has a perfect view over the area in which our hydrophones are deployed.  From the eastern point of Strawberry Island, you can see humpback whales and harbor seals, as well as Stellar sea lions, harbor porpoises, and sea otters.  And off in the distance, two black bears comb the rocky shores to the south of our camp.  The salmonberries are ripe, the forest is dense, and the views are incredible. 


The orange star is the location of our camp, the green markers are the locations of our hydrophones

Strawberry Island has a pretty stellar backdrop.  (Photo: L. Matthews)

We’ve just returned from the first of four stints on the island.  Each stint lasts around eighteen days, then there’s a four-day break off the island in the local town of Gustavus (real showers, cell phone service, and the best Rueben money can buy).  Overall we’re off to a great start – the entire field team is trained on all the protocols and we’ve already collected some great data.  Excited to see what’s to come in stint number two!


FUN SCIENCE FACT #43:  Humpback whales have the potential to lose up to 1/3 of their body weight during the winter.  When considering that humpback whales upwards of 70,000 pounds, that's a pretty impressive number of pounds to lose.

Humpback whale in Glacier Bay (Photo: L. Matthews)