Monday, September 17, 2007

West Ice survey 2002:IV

Read the first part of this post here
Those forays on the floes, although enormous fun, were really an aside. We had to be out on the ice for a few weeks, but only needed a few days for the aerial surveys. So tagging was useful, but not the primary objective of the expedition. Our main task was to fly over the whelping patches of females and count newborn pups.

The harp seals that meet off east Greenland are wanderers of the Arctic Atlantic. After giving birth in the West Ice, they either swim north along Greenland's east coast, east to Svalbard; or to Iceland and points south - some almost to Scotland's Western Isles - with occasional forays to the Norwegian coast. The only time all members of the population seem to get together is at pupping. Even then, the males are mostly underwater, females often in the water, and only the pups stay on ice floes.

This means that pups are the most reliable animals to count. That was why we had the helicopter, to fly transects - regularly spaced lines - over each patch in order to estimate the number of pups born. But there's a catch (in field biology, there's always a catch). All harp seals in a population give birth over a short period in March and April - exactly when varies between the three separate Arctic populations. Generally, the females in an individual whelping patch all give birth within two weeks of each other. Harp pups suckle for less than two weeks - the average is 12 days. Towards the end of those 12 days, pups moult their trademark white fur, replace it with their adult coat, and start contemplating their future as swimmers, rather than ice floe couch-potatoes. This means that once pups have moulted their white fur, there's no guarantee that they'll stay on the ice. Getting the survey timing right – including allowing for weather and pups' maturation – is the crucial balancing act of the exercise.

While they're suckling, pups change size and shape so quickly that it's possible to estimate their age to within a couple of days. Seal biologists have developed a system for categorizing pups into seven stages of these first few days of life, based on how they look. So, by flying a series of surveys using the helicopter, in which we estimate the stages of each pup seen, we could get a quick-and-dirty overview of the relative age of the pups there. Additionally, we recorded the stage of every pup that we tagged when we were out on the ice. Between the “staging surveys” as Garry called them (they're a Canadian invention), and our tagging data we had a pretty good sense of the age of pups in a patch.

Initially, we'd see many tiny white pups, lots of fresh-frozen meaty afterbirths, and some very satisfied looking gulls. After a few days, newborns became less common. We had to make some tradeoffs with timing for running a counting survey - flying over the patch to estimate the number of pups born, as distinct from the staging surveys – too early, and there'd be pregnant females with pups still to be born, too late and the some pups could have already taken to the water. And of course, we needed to think about weather and safety, too.

The Arctic's like that.

It was time to prepare for the counting survey, to work out our observers' strip widths.

We needed to know the area covered by each transect – the length was easily calculated from GPS-derived start and stop points, but width is slightly more complex. Strip width, from simple geometry, depends on the height that the helicopter flies – higher gives a wider strip, but fly too high and pups get hard to see. This is the most simple form of what's known in the trade as distance sampling - the observers should be able to reliably count every pup that is within the strip, and so there to be counted. If the area through the helicopter window was too large, pups would be missed.

Once our strip width was established, we'd know the area (transect length multiplied by strip width) that we'd sampled. As we also knew the total area occupied by all the seals in the whelping patch, we could calculate the proportion of the patch that we'd sampled. From years of experience, the Canadians worked with a strip of 30 metres when the helicopter flew at 100 feet.

Ivan and I were the observers. We needed to get our 30 metre strips marked out. So, Ivan Tore, Garry and I jumped in the helicopter and flew out to some large pans well away from any seals. We checked there were no polar bears around (as we hadn't brought rifles), then landed. Tore and Garry got out with a measuring tape. They marked out two strips, 30 metres apart, then lay inside them, doing seal impersonations. Ivan and I were taken to 100 feet above the ice - the helicopter had an altimeter that used a laser for measurement, so we knew it was accurate. Then we made ourselves as comfortable as possible while turned sideways in a helicopter seat, and nursing an old laptop computer. One at a time, we directed the pilot until the lower strip on the ground was just blocked by the helicopter's lower skid, then marked off the top line with a bit of electrical tape on the window. Field biology survives on electrical tape, duct tape and cable ties. We retrieved our seal impersonators and flew back to Lance.

And then Garry, Ivan and I practised. Our laptops were linked to a GPS receiver, with a little program (courtesy of one of Garry's computer wizards) written so that whenever we pressed the keyboard, it stored a position fix from the GPS. Unfortunately, it only worked on Windows 98 machines, so we'd had to drag out a couple of somewhat antique laptops for the trip. Ivan and I had to tap a key to record each time we saw a harp seal pup. As well, we'd call out our sightings as we made them, into the microphones of the aircraft headsets that we all wore. Garry recorded our numbers into a notebook as backup to the computers.

All my previous experience with strip transects had been with whales, dolphins or dugongs1 – mammals in the ocean, with an unfortunate habit of diving, and so being mostly out of view. To me, counting animals that lay stationary on the ice seemed a bit like cheating. A few practice runs and we were ready for the real thing.

These visual surveys from the helicopter don't provide a permanent record of the seals we actually saw. So, Kjell and his crew in the plane were also going to fly over each patch. As the plane had an aerial photogrammetric camera aboard, they could photograph a series of transects over the whelping patches as well. Photographs provide a permanent record of the survey. The concept is exactly the same as the helicopter counts, except that the photos create their own strips.

As the time neared to run the full survey on the first patch we found, Kjell radioed to announce that from the plane they'd found another, far larger patch a few hours' steam to the west. Lance's engines came back to life for a few hours.

Read the next part of this post here.

No comments: