The Secret Lives of Southern Giant Petrels

Thursday 14 June 2018, 11:30pm–12:30am

This talk is presented by Barbara Wienecke (Australian Antarctic Division) on behalf of her co-workers John van den Hoff (AAD), John McKinlay (AAD), Sasa Otovic (UTas) and Madalyn Riley (UTas). Barbara is a seabird ecologist who has studied mainly penguins over the years. But the flying birds are equally fascinating! This seminar focuses on a study of Southern Giant Petrels that was based on images obtained from automated cameras that operated all year long.


Southern Giant Petrels (SGPs) are the largest flying birds that breed in Antarctica. Although they have a circumpolar distribution, only four small colonies are known in East Antarctica; three of those are in the Australian Antarctic Territory and all are located in Antarctic Specially Protected Areas (ASPAs). Each ASPA has a management that is reviewed every few years. Since these birds are easily disturbed by the presence of humans, we are trying to find non-intrusive ways to obtain data that help to determine whether the management plans are successful in protecting the values of the Areas. This talk reports on the efforts to test the utility of automated camera systems — already successfully deployed near Adélie penguin colonies — to collect data on the phenology of SGPs and the potential to monitor these birds long term with minimal human impact in a cost effective way.

Automated cameras were deployed at Hawker Island (near Davis station) and at Nelly Island (near Casey station). We analysed of information from images obtained over a three-year period and compared our findings between the two islands. The cameras allow us to gain interesting insights into the behaviours of SGPs and deliver some beautiful images as well!

Please join us in the theatrette of the Australian Antarctic Division. All welcome!

AAD Seminar Team

The secret lives of Southern Giant Petrels

Video transcript

Lauren: Welcome everyone. We might get started with today's seminar which will be given by Dr Barbara Wienecke who is a senior research scientist at the Australian Antarctic Division. Barb has been here at the division working since 1993. She has spent time on various sub-Antarctic islands and at non-Australian research stations, as well as our own research stations. Barb has studied a number of different penguin species from Little penguins all the way through to Emperor penguins but also some flying birds, such as shearwaters, Black-browed albatross and those she will be talking about today, the Southern giant petrels. I hand over to Barb to give her presentation.

Barbara: Thank you very much, Lauren, and thank you very much for coming. It's always lovely to see so many friendly faces and potential fellow giant petrel lovers. Before I start I would like to acknowledge my co-authors for this talk; John van den Hoff and Ian Hay conceived this project nearly 10 yours ago now - you wouldn't believe it... John McKinlay thank goodness came to our rescue with his statistical expertise, and a lot of the hard, heavy lifting and hackwork was done by two students, Sasa Otovic and Madalyn Riley.

Now a brief outline: I'll introduce you to the giant petrels - one should never assume that everyone knows everybody; I'll give you a brief outline of how we went about carrying out this project, and I shall also give you some results.

Giant petrels belong to the genus Macronectes. They are restricted entirely to the southern hemisphere. Giant petrels are also the biggest of the fulmarine petrels by far. The fulmarines include the Snow petrels and Antarctic petrels so they are certainly an order of magnitude in size larger than those little guys. Two absolute heroes in the ornithological world, Bill Bourne and John Warham, discovered that - while working on SGPs at Macquarie Island - that we actually have two species. That was in 1966 and I think it's very special that this particular discovery was made at Macquarie Island.

So these days we talk about Northern and Southern GPs and when you look at them you think: Really? What on earth is the difference? But as with so many things in ecology you have to look a little bit more closely. One thing you notice when you focus on the tip of the beak is that in Northern giant petrels it is this lovely rusty red, and in Southern giant petrels a most delightful green. But other than that there are also differences in their breeding phenology - that is the breeding cycle; there are size differences and morphometrics. SGPs are a little bit bigger than - and there are also plumage colouration differences.

It is only the white phase that occurs amongst the Southern giant petrels and they are just so beautiful. They are always spectacular to see.

Both of them are extraordinary scavengers and you only need to look at the beak to see that they definetely evolved some good flesh taring capabilities. But of course if there is no carrion, they are also exquisite hunters and one tends to forget that sometimes.

They used to be very badly affected by longline fisheries. As scavengers they are of course attracted to offal. As petrels they have an extraordinary sense of smell and unfortunately can pick up the scent over great distances. But thankfully through massively improved fisheries management strategies, particularly through CCAMLR [...] that trend seems to have been reversed.

In terms of their distribution, it is interesting to see that there is really quite a difference as well. The red triangles show you the breeding locations of Southern giant petrels and I hope you can see a handful of purple ones. Those are the islands where Northern giant petrels exclusively breed. The green boxes indicate where both these species occur.

As far as their population is concerned, again there are quite significant differences. Most of them occupy areas in the southern Atlantic Ocean and the four little arrows indicate the breeding locations that we know of on the Antarctic continent in the east. You can also see the size of circle roughly gives you an indication of the size of the breeding population. The colours of the dots indicate the population trend. You see one large blob of red right next to South America where there is an apparent increase. This large blob is the Falkland Islands...

... absolutely the capital city for Southern giant petrels. A survey in 2004 indicated that there are roughly 19,500 breeding pairs. And the amazing thing is the arrow points to a very special island... of the two Elephant Cays of the Falkland Islands where in this small area encircled apparently approximately 11,000 of them breed. Now I find this just stunning. That might be worth a trip to the Falklands just to experience that.

Population estimates are a really tricky thing to obtain. For many colonies there simply are no current data - we're talking globally. Some of the data are 70 to 80 years old. When we read studies and publications we find that there are often differences in count units, there are differences in census timing and there are differences in census methods.

So using a highly non-standardised approach is making any interpretation really difficult.

But when you at these data that were taken from Mike Dunn from BAS {...} he published this a couple of years ago, it shows you how much giant petrel populations can fluctuate. On the y-axis you have the number of individuals, on the x-axis the years, the black dots indicate the number of breeding pairs and the brown ones are the chicks. You notice that there are a couple of gaps. When you send people into the field sometimes things go pear shape. Due to weather or logistics you can't get there. So often data sets are incomplete. You also notice that in 2003 although a lot of birds attended the colony and attempted to breed, the number of chicks was really quite low. So a large number of adults in the colony does not necessarily equate to high breeding success. The other thing that is quite noticeable is that there can be a massive change in numbers attending the colony from one year to the next.

We are all aware that things are changing. Just as an example, and an example that is dear to our hearts, the Antarctic Sea Ice extent. This is a summary from 2002 to 2017, and - that's a little difficult to see in this spaghetti knot - but you can possibly make out that the blues and green curves are fairly close together in the centre, and then from about 2012 onwards the sea ice started to expand and NASA - for several years in a row - said this is the maximal extent ever reported. O, actually THIS year is the largest extent we've ever seen. But then from 2016 things went exactly the reverse and we now talk about the most minimal extent ever recorded.

I had a quick look and unfortunately even this year we are tracking well below average. Now every shift in this curve of course is related to millions of square kilometres of change that is happening.

Do environmental changes impact SPGs? In order to answer this we really need a very good understanding of what drives this population. You would think that by now we have a fairly good understanding of the basic biology but it goes beyond that. We know that differences occur at different islands but we simply cannot extrapolate from island to the next. To answer our questions we really need valuable and detailed observations, preferably collected on an annual basis.

We had two study sites for this work: Hawker Island near Davis, and the Frazier Islands near Casey. Both of these sites have been Antarctic Specially Protected Areas for a very long time designed to protect the breeding areas at the southernmost extent of the birds' breeding range.

A lovely picture of Hawker Islands! It's fairly flat and ice free and this red circle shows you the location of the one and only location of giant petrels on the island. So they are certainly not limited by habitat

The Frazier Islands are a group of three small islands about 13 or 14 nautical miles off Casey. All three islands are inhabited by southern giant petrels but we only monitored those present at Nelly Island.

To give you a bit of an idea of what we actually knew about these colonies, I use Hawker Island as an example, most of the information we have for these islands was collected from the 1970s to the 1986. Here too we found that the count units varied; there are reports either from adults present or occupied nests, sometimes even just nest, or the number of chicks. Even the timing of counts was highly variable. Adults were counted any time in November, December or January and you very quickly start to get the idea that we might be comparing apples with pears if we treat all these data equally. Chick counts were equally variable and ranged in time from January to as late April.

Why did we undertake this work? As you have figured, East Antarctic colonies have never really been studied in detail. Both study sites are Antarctic Specially protected Areas which means they come with a management plan and our lovely policy colleagues come to us regularly to update it and we no new information to provide them which is really a tad embarrassing. The other thing is that Southern giant petrels are also and ACAP {...} listed species so we are responsible.

We learned a wonderful lesson from our colleagues in the Adelie penguin world. We decided to use automated camera deployments. Colin already talked this in previous years and highlighted how successful they were for monitoring Adelie penguin populations and our very own Kym Newbery had really done a fantastic job setting this system up. So we figured we see whether we can use this on giant petrels as well.

But rather than just playing a bit in the field we decided we better ask some specific questions and we eased ourselves into it with an easy one: we just wanted to find out whether SPGs are present in winter. The cameras were set up to operate all year round. We also wondered whether we could identify events in the breeding cycle based on the images that we had. And as I mentioned before, one should never extrapolate from island to the next. We deliberately chose two islands bout 1500 k apart to see whether we could spot any differences. So what we actually look for were identifiable events in the annual cycle.

We did our work before we started. We did a bit of a literature search. One very important date in any bird's life is the lay date. Reproduction is what the game is all about and you want to get it right. Don't worry too much about this lengthy list but what I try to highlight is here is again an inconsistency in the way data are either collected or published. Here is an example where just a very broad range of lay dates was given; in other studies at least we had a date range and a few studies thankfully gave us actually a mean and a standard error attached to it. When you just have quick look at the timing you can see they could have been as early as Augusts and all the way into November. So from a latitudinal gradient from about 40 degrees to 68 degrees s outh there is quite a difference in timing.

We also thought we better educate ourselves about behaviours. It just seemed a good idea, something that could come in handy. Thankfully we found a paper by Vincent Bretagnolle, a French researcher who did fantastic work on SGPs at the Crozet Islands, and he wrote this marvellous paper about the social behaviours. Not only did he write a paper; he also provided us with fabulous little drawings and that was very exciting. We thought how difficult can it be to identify any actions in the images. Can't be a problem. Well, one should never celebrate too early...

... after all we're dealing with grey-brown birds on grey-brown rocks. Fabulous! So just to help you out there are actual occupied nests. And when you think well in summer that's a bit of a pain but surely you can see them in winter...

... even on snow, they still look like rocks!

Another minor problem we had with one of the cameras that happened to face east is sun flare. But when the camera point directly into the sun of course you do not get a great deal out of that wonderful sunrise. This is still a pretty good picture but you get the idea. It can be difficult to identify things. However, at this time of year we took several photos per day and we felt we didn't lose out on any information. We just got extra pretty pictures.

The other challenge of course was the weather. There is always weather and despite Kym's best efforts to keep the cameras going, every now and then there were blizzards that were severe and clogged up the lens sometimes for a couple of days so we couldn't see anything.

How did we go about this? That was a bit tricky. We had to work a little bit backwards. We wanted to identify specific events, and we needed to make sure that we had active nests. We took a few images that were taken by the cameras in February in a given year and simply circled all the nests with big fluffy chicks in them and we numbered them. Then we went to the following year and did exactly the same and sometimes the same nests were occupied and other times more nests were occupied. We kept the numbering consistent and we excluded the nests which were not all that clearly visible. We really needed to see what was going on and we couldn't afford to worry about nests that clearly were active but were hidden behind rocks.

We recorded the dates of events and some of these events were quite colony specific that means that for example we just recorded the date when we saw the very first adult in the colony. It didn't matter whether that bird was sitting on a nest or just walking around. But some of the more important variables were nest specific. We looked at the timing of pair formation; we tried to figure out when eggs were laid and when the chicks hatched, and when they became independent. Independence here refers to the chick's ability to regulate its own body temperature. It all seemed to be fairly straight forward, doesn't it?

Yes, some variables were quite easy to identify. Timing of pair formation, fantastic! All you need to do is look at the series of images until you find one where two birds are sitting on the nest. This doesn't necessarily mean that they are a pair but when they sit there for a couple of days and keep returning you can be fairly certain that something is going on.

Other dates were far harder to identify and one of them was the lay date. It took us a while to realise that we actually had to look for really quite subtle changes in body posture. So we really had to put on our detective hat and start looking very carefully. I'll show you a couple of examples.

Here is a giant petrel incubating. Its partner had left and the bird was sitting on the nest for days on end so we were pretty confident that this was an active nest. Look at the posture of the bird: the wings are nicely, tightly folded against its body, the wing tips are folded over the feathers of the tail. A nice compact bird. Literally, the second that a chick hatches the entire posture changes. The bird seems to melt into the rocks, the wings are more spread out and you can just about feel a little guy crawling underneath mum or dad. This was really not an exceptional event; we saw it at pretty much all the nests. In some nests it was a little bit harder to see than in others but just to give you another example, here again a nice compact bird and look at this fabulous pancake shape it has taken to here. Just great!

The reaching of homeothermy or thermal independence is of course a gradual process. It was actually quite tricky to put a date onto it. You can see the little chick there. It is still mainly covered by the parent so we would not have declared this as thermally independent.

It takes about two to three weeks and finally there is a day when we see chick quite happily sitting next to mum or dad by itself. Mum or Dad is still around. We are now in the guard phase; in case of bad weather the adults will again hop onto their chick to protect it. But increasingly more the chicks are being left to their own devices and spend longer and longer periods by themselves in the nest. Lovely... they are just fabulous birds!

So the end of the guard phase is when the chick sits by itself on the nest and thankfully these little guys are beginning to grow and even though they are still grey - maybe our eyes had tuned into the images - they seemed to be a little bit easier to identify.

As the chicks approach fledging they become really quite mobile. It was remarkable to see - although the terrain is quite open - how the chicks simply remained at the nest. Even if the parents were away for a couple of days these chicks simply did not move. Sun or snow, they stayed put. But as they approached fledging - that means they shed their down and the adult feathers are growing - they really need to start running around a little bit and start exercising those muscles. The little chap in the centre shows how difficult it is to spread these very long wings and maybe even fold them back together again.

The last chick in the colony - that was one of those things that was a colony specific event - that could be quite difficult to determine. We are a little bit cautious about this simply because what the camera shows is of course only a certain area of the island. There could be a chick or an adult sitting right next to the camera and we would never know about it. But we figured just for completion sake we might as well note these dates down.

Are the automated cameras delivering? I would answer this with a loud and resounding YES. There were a few minor disadvantages but without any doubt they are far outweighed by the many, many advantages that these systems offer. There is no human disturbance to the birds. The cameras at Hawker Island were downloaded and serviced outside the breeding season; the cameras at the Fraziers were actually left out there for four years and there was only one camera that failed after 449 days. That was just a remarkable outcome! The wonderful thing also is that we do get year round information, literally we have a bit of information for nearly every day of the year. Another huge advantage is the archival nature of the data. We have images we that used for a particular purpose but they can be used to answer questions as well. It didn't take long to figure out that one can answer questions related not just to phenology but a to great variety of questions and topics.

Are SGPs around all year? Yes. That was really a surprise. For some reason a lot of us assumed that they weren't. Why on earth would you hang around in winter? But they did. The other interesting thing was that the birds we saw in those images were - regardless of the amount of snow in the colony - they were sitting on known nest sites. So it was not just visitors that accidentally came through. If you think about the picture of Hawker Island, it would really be extraordinary and coincidental if any accidentally passing giant petrel happened to sit on a known nest. Because that happened day in and day out. There was a period in May and June when the birds were rarest. There were several days that we didn't see anybody but they literally are present all year round.

Could we identify the key events? And again I am happy to report that we can answer this question with Yes. This table summarizes most of our findings. On the left hand side you see the various variables that we examined and it is a plot over time. So we go from June until May. The green blobs are the data from Hawker Island and the blues are Frazier or Nelly Island. You can see that at the beginning of the season - or what we declared to be the beginning of the season - up to pair formation the dates are incredibly variable. There was a date within and between colonies of 31 days when the first pair formed at any one nest. But as soon as you get into the breeding season you can see how unexpectedly and highly synchronized these birds actually are. It gets a bit more fluffy again towards the end of the season but could in part be due simply to the restricted view of the cameras.

Just summarizing in a slightly other way. When we compare the islands... Hawker Island in general seems to be a little bit later in many of the events that we recorded. But laying was extraordinary. It is the 26th or 27th of October, that's when you lay. That was across three years. Now we have more data from Hawker Island and it'll be very interesting to see whether this holds true for an extended period.

There is of course more work we would love to do. Another advantage of the cameras is that they record things like storm events. You can't see anything, there must be a blizzard. That conclusion is fairly straight forward. We can go through the images and see how many storm events there were per season, we can figure out how long they were. We are trying to figure out with John McKinlay's help whether we can roughly estimate the amount of snow that has fallen but that is very complicated. But we can certainly try and relate these findings to breeding success. We can also look at individual nests over time series and see how frequently a particular nest is occupied because giant petrels don't necessarily breed every year. We found that oftentimes there are birds hopping around. They are sitting on the nest sometimes for up to a week, and then they fly away and there is nothing there. Just seeing birds on the ground does not mean that they are actually breeding pairs which makes any population count extremely difficult. We are even contemplating whether we should attempt to estimate foraging trip durations because in some pairs at least the partners are so differently coloured that we can probably identify them and distinguish them from each other. That would of course be hugely exciting if that would work.

When looking through these images one can help but be reminded that life really isn't always that easy even for birds that are quite clearly exceptionally well adapted to deal and live in environments that we experience as quite extreme. But it doesn't mean that the birds are doing it particularly easy themselves. I just draw your attention to again the chick in the centre and for future reference the rock on the right hand side. So here we are on a not too crash hot day in January 2012. It is just after noon and the chick is still happily parked up next to mum or dad.

A few hours later, the rock is still there but where is the chick? It was completely and entirely covered by snow. The parent had moved on. That is fair enough. It took until one o'clock the following morning before a parent returned. It was interesting; not only is it sitting where the chick presumably was. You can also see that the snow around this area was heavily disturbed. The only conclusion one can draw is that the parent was trying to reach its chick. In this picture you can also see that adult was vocalising and probably calling its chick.

Here we are four hours later. The bird is still at it but still no sign of the little fluff-ball.

And finally, even later, the adult has now moved aside. Obviously it feels it has done the work and the little body is exposed but is it actually still alive? It was really amazing. It took quite a few more hours - quite a few more images - before we could answer this.

It was absolutely delightful nearly a day and half later that the little pumpkin was alive and feeding! This sort of little incidences really make your day.

There was one thing we never knew about giant petrels - and again that's why cameras are so fantastic -  cameras take images without the birds knowing - we thought

... They like to take selfies! It was so fabulous. We never saw any sign that would be indicating that the birds were bothered by these cameras but some of them were quite clearly attracted by this little zooming sound that was made when the cover of the lens moved aside before it took a picture.

And it was not just one! and you can also see that some of them are clearly ...

... better at it than others.

Giant petrels really are amazingly complicated birds. Their breeding cycle is far more complex than we ever expected. We really hope that we can utilize this information to try and determine times when it is most sensible to perform counts, especially of adults. As I said before just because there are birds sitting in the colony does not necessarily mean that you have breeding pairs. That is my biggest concern about any of the publications, the censuses of giant petrels. People tend to count whoever is there and then report it as breeding pairs. But there is possibly a world of difference between those two and that is just something to be aware of. I hope that you share my enthusiasm for giant petrels and if there are any questions...

... I am more than happy to answer them. Thank you very much!

[end transcript]