Archive | September 2013

Certainty score

Back in June, I wrote about algorithms I was working on to take the volunteer data and spit out the “correct” classification of for each image. First, I made a simple majority-rules algorithm and compared its results to several thousand classifications done by experts. Then, when the algorithm came up with no answer for some of the images (because there were no answers in the majority), I tried a plurality algorithm, which just looked to see which species got the most votes, even if it didn’t get more than half the votes. It worked well, so I’m using the plurality algorithm going forward.

One of the things I’ve been curious about is whether we can detect when particular images are “hard.” You know what I mean by hard: animals smack up in front of the camera lens, animals way back on the horizon, animals with just a tip of the ear or a tuft of tail peeking onto the image from one side, animals obfuscated by trees or the dark of night.

So how can we judge “hard”? One way is to look at the “evenness” of the volunteer votes. Luckily, in ecology, we deal with evenness a lot. We frequently want to know what species are present in a given area. But we also want to know more than that. We want to know if some species are very dominant in that area or if species are fairly evenly distributed. For example, in a famous agricultural ecology paper*, Cornell entomologist Richard Root found that insect herbivore (pest) species on collard greens were less even on collards grown in a big plot with only other collards around versus on those grown in a row surrounded by meadow plants. In other words, the insect species in the big plot were skewed toward many individuals of just a few species, whereas in the the meadow rows, there were a lot more species with fewer individuals of each species.

We can adopt a species evenness metric called “Pielou’s evenness index” (which, for you information theorists, is closely related to Shannon entropy.)

[An aside: I was surprised to learn that this index is named for a woman: Dr. Evelyn Chrystalla Pielou. Upon reflection, this is the first time in my 22 years of formal education (in math, computer science, and ecology) that I have come across a mathematical term named for a woman. Jacqueline Gill, who writes a great paleo-ecology blog, has a nice piece honoring Dr. Pielou and her accomplishments.]

Okay, back to the Pielou index: we can use it to judge how even the votes are. If all the votes are for the same species, we can have high confidence. But if we have 3 votes for elephant and 3 votes for rhino and 3 votes for wildebeest and 3 votes for hippo, then we have very low confidence. The way the Pielou index works out, a 0 means all the votes are for the same species (high skew, high confidence) and a 1 means there are at least two species and they all got the same number of votes (high evenness, low confidence). Numbers in between 0 and 1 are somewhere between highly skewed (e.g. 0.2) and really even (e.g. 0.9).

Another way we could measure the difficulty of an image is to look at how many people click “nothing here.” I don’t like it, but I suspect that some people use “nothing here” as an “I don’t know” button. Alternatively, if animals are really far away, “nothing here” is a reasonable choice. We might assume that the percentage of “nothing here” votes correlates with the difficulty of the image.

I calculated the Pielou evenness index (after excluding “nothing here” votes) and the fraction of “nothing here” votes for the single-species images that were classified by experts. And then I plotted them. Here I have the Pielou index on the x-axis and the fraction of “nothing here” votes on the y-axis. The small pink dots are the 3,775 images that the algorithm and the experts agreed on, the big blue dots are the 84 images that the plurality algorithm got wrong, and the open circles are the 29 images that the experts marked as “impossible.”  (Click to enlarge.)

Pielou-and-blanksAnd sure enough, we see that the images the algorithm got wrong had relatively high Pielou scores. And the images that were “impossible” had either high Pielou scores or a high fraction of “nothing here” votes (or both). I checked out the four anomalies over on the left with a Pielou score of zero. All four were unanimously voted as wildebeest. For the three “impossibles,” both Ali and I agree that wildebeest is a reasonable answer. But Ali contends that the image the algorithm got wrong is almost certainly a buffalo. (It IS a hard image, though — right up near the camera, and at night.)

So we do seem to be able to get an idea of which images are hardest. But note that there are a lot more correct answers with high Pielou scores and high “nothing here” fractions than errors or “impossibles”. We don’t want to throw out good data, so we can’t just ignore the high-scorers. But we can attach a measure of certainty to each of our algorithm’s answers.

* Richard B. Root. 1973. Organization of a Plant-Arthropod Association in Simple and Diverse Habitats: The Fauna of Collards (Brassica oleracea). Ecological Monographs, Vol. 43, No. 1, pp. 95-124.

Advertisements

The best view is from a balloon

Don’t get me wrong – it’s really nice to have running water, internet, and my pick of fresh vegetables from the weekend farmer’s market. But sometimes I miss the Serengeti. Watching from my window in Minnesota, I’m lucky if I see a squirrel. And let’s face it — squirrels are only so exciting, for so long.

One of my favorite ways to see the Serengeti was from the unbeatable vantage point of a hot air balloon. Yes, that’s right. We might not have an indoor toilet or fresh food, but we have hot air balloons. Okay, that’s not entirely correct: Serengeti Balloon Safaris has hot air balloons that fly tourists over the heart of the park. In fact, you’ve probably seen them floating past in the camera trap photos:

IMAG0255

SBS has been flying balloons in Serengeti since the 1980’s, and has always helped us researchers out whenever they could, from letting us drag our hand-held tracking equipment up for flight to listen for lost lion prides, to letting us tinker with our Landrovers in their garage. In fact, they sponsored a lion cub during one of our fundraising campaigns a few years back – and now there’s a cub named “Balloo” living in the Mukoma Gypsies pride in the heart of Serengeti.

I went up for my first flight in 2010. George Lohay, Stan’s predecessor on the project, and I had to wake up at 4 am to make it to the launch field on time. That really is as terrible as it sounds, however, that morning it saved us from an invasion by the relentless carnivorous safari ants (siafu). Well, to be more exact, we were able to flee the house before the ants had invaded our beds, meaning we escaped with minimal damage. And by the time we returned that afternoon, the ants had already moved on.

IMG_0707 IMG_0742 IMG_0800 IMG_0775

I’ve been thinking about balloons lately because one of SBS’s pilots and a dear friend of the Serengeti Lion Project & Snapshot Serengeti, Jason Adams, is currently preparing to defend his title in Canada’s upcoming National Hot Air Ballooning Championships.  Everyone in Serengeti and on the Snapshot team will be rooting for him.  Good luck Jay!

Top speed: Technology, movement, and the cheetah’s secret weapon

I got to spend all of last week at a movement ecology workshop in Zurich, Switzerland – conveniently beating the heat wave Minnesota has apparently been having!

Migration patterns of the sooty shearwater, revealed by Scott Shaffer of UC Santa Cruz in a new study.

Movement ecology explores both how and why the way animals move the way they do, and what this means for them as individuals, as populations, and species. What triggers do animals use to decide where to go and what to do while they’re there? Why are some species territorial, while others overlap? Why are some species migratory? What do these behaviors mean for their individual fitness, their population dynamics, and global distributions? How does our understanding of animal movement change the way we try to protect species and the habitats they need?

It’s a pretty big field…and it’s one that is growing in leaps and bounds with modern technology – from camera traps to GPS collars, vast satellite networks and high resolution global imaging – we are able to ask questions about movement and distribution that would have been impossible just a few decades ago.

_MG_5710

For example, researchers from London’s Royal Veterinary College recently discovered that cheetahs – contrary to popular opinion – might not rely on their top speed to catch their dinner, but instead their agility and acceleration. In captivity, cheetahs had been clocked at nearly 105 km/hour – but new, lightweight, super-fancy GPS collars let researchers take these speed tests to the wild. What they found was that cheetahs in the wild averaged about half that speed – but that they accelerated and turned with unparalleled power and agility. Which means that our long-standing perception of cheetahs as needing wide open spaces for really fast chases might not be accurate – and that they might actually be able to hunt quite well in the woods!

This is just one of many, many secrets that the field of movement ecology, with the help of modern GPS technology, is uncovering.  Some other stories are a little less exotic and closer to home. The Today Show recently covered the amusing tale of a concerned cat owner who designed a tiny tracking device to see just where her pet was going every day. Her sleuthing turned up some interesting results – not just for cat owners, but for scientists too — and scientists are now encouraging cat owners around the country to track their cats and share their data on an online repository called Movebank.org.  So if you have a cat and want to see where it goes when you’re not looking, or help scientists understand how domestic cats fit in to the larger ecosystem, check it out!

March of the Elephants

When you think of elephants, you may immediately think of their defining characteristics: trunks, big ears, tusks. Or you may think about how they live in large family groups and are very social. You may even think about the story of the blind men and the elephant. You probably don’t think about them as engineers of their ecosystem. But they are.

Elephants are native to the Serengeti ecosystem, but Serengeti elephants were likely all killed off for ivory in the 1800’s. At least, there weren’t any recorded there until the middle of the twentieth century when they started moving back in again. In the 1960’s they migrated in from both the north and the south, and by 1970 there were over 3,000 elephants in the Serengeti. Things got rocky for elephants again in the 1980’s as severe poaching reduced their numbers in Serengeti National Park to around 500. In 1988, elephants were given CITES endangered species status and worldwide trade in ivory was banned. This was good news for Serengeti elephants and their numbers rebounded again into the thousands.

These ups and downs in elephant population have allowed scientists to study the impact elephants have on their environment. I’ve written before about how the rainfall patterns in the Serengeti affect grasses, and about the role that fire plays. Elephants have their greatest impact on trees. Elephants eat both grasses and trees, but depend on trees for food during the dry season.

In the first half of the twentieth century, the number of trees per hectare was slowly declining across the Serengeti. But starting in the 1970’s, the number of trees rapidly increased. Scientists believe that the initial decrease in trees was due to the the disease rinderpest. Rinderpest killed off the majority of Serengeti’s wildebeest, allowing the grass to grow tall, and fueling huge, strong fires. These fires killed most tree seedlings, meaning that as trees died, they were not being replaced. When rinderpest was halted, the wildebeest population exploded, and the wildebeest kept the grass short and the fires smaller, allowing trees to gain a foothold once more.

Okay, but what about elephants? Well, elephants eat trees — especially small, tender saplings and regrowth from trees damaged by fire. In the 1980’s, while poaching was particularly severe on the Tanzanian side of Serengeti (Serengeti National Park), the Kenyan part of Serengeti (Maasai Mara) mounted a strong anti-poaching effort and kept its elephant numbers high. Across the Serengeti, the trees were increasing, but in the Maasai Mara there were also a lot of elephants. It turns out that in the Maasai Mara, the trees didn’t increase like they did across the border in Tanzania where the elephants had been greatly reduced. Instead the high number of elephants in the Maasai Mara is keeping tree numbers down, despite the reduction in fire intensity.

So elephants are key players in maintaining what scientists call “alternative stable states” in the Serengeti. While there are plenty of elephants once again in the Tanzanian part of the Serengeti, they don’t reduce the higher tree numbers; only fire can do that. But on the Kenyan side of the border, tree numbers remain low because elephants there have been continuously eating saplings; the tree population cannot increase because of the constant elephant pressure. The key difference between the two areas is simply their history.

I think the fourth blind man should get extra credit.

The Fourth reached out an eager hand,
And felt about the knee
“What most this wondrous beast is like
Is mighty plain,” quoth he:
“‘Tis clear enough the Elephant
Is very like a TREE!”