If you’ve been clicking through our new “extended survey” season, you might have noticed some new critters — particularly lots of cattle, sheep, goats, even donkeys! This is because the extended survey reaches much farther north, west, south, and east than the long-term camera study area. You can see all of Tom’s and Michael’s cameras on the map below in the black and gray circles. Two sites are at the edge of the park, where we might expect to see the occasional pastoralist passing by (even if they’re not supposed to) — and one site is *all* the way down deep in the Ngorongoro Conservation Area, where pastoralists live permanently. If you see any livestock in the images, go ahead and mark them as “cattle” — we’ll be able to figure it out from there!
Join us for a special edition of Snapshot Serengeti! Hot on the heels of Season 9, we’re launching the Snapshot Serengeti Extended Survey – a season loaded with photos from new locations in the Serengeti Ecosystem.
The original Snapshot Serengeti team has joined up with Prof. Michael Anderson (from Wake Forest University), Prof. Rico Holdo (from University of Georgia) and Dr. Tom Morrison (from University of Glasgow, UK) to understand how wildebeest, zebra and other herbivores impact vegetation dynamics in Serengeti.
We are particularly interested in understanding how the many many herbivores in Serengeti (both in terms of total number and the diversity of species) impact the growth, survival and germination of savanna trees. But why study trees? Savannas – like the one found throughout the Serengeti – present a bit of a paradox for ecologists. Most savannas receive sufficient rainfall throughout the year that they “should” become forests. We know that things like fire and herbivory help them maintain a mixture of continuous grasslands interspersed with smaller number of trees. Indeed, Serengeti receives enough rainfall (450mm per year at the driest part of the ecosystem to over 1000mm in the wettest) that it should be chocked full of trees, but yet, for most of the ecosystem, it remains a savanna (and quite a beautiful one).
We’ve designed an experiment understand the fate (the life and death and growth) of small trees at the seedling and sapling stage. During the wet season, we germinated a large number of seedlings in a nursery at the research center in Serengeti (see photo). We selected the two most dominant trees in the ecosystem: Acacia tortilis (commonly known as the umbrella acacia) and Acacia robusta (the stink bark acacia).
After growing 760 seedlings for about 6-8 weeks in the nursery, we transplanted them in the field at 19 different plots across the ecosystem, spanning the large gradient in rainfall mentioned above. Once in the ground, we subject them to various treatments (or mistreatments, if you like) such as fire, herbivory and watering. By using a combination of these different treatments, we hope to understand how important they are in determining the survival of trees at a small stage, which will ultimately inform why they survival into adulthood.
Our camera traps are set up at all of these different plots so we can track how herbivores use these areas differently and the effect that has on the vegetation, and we need your help to classify the photos! So head on over to Snapshot Serengeti and dive in! You never know what you might find!
It’s finally out! 9 long months ago, we received the good news that our second Snapshot Serengeti paper was accepted for publication in Conservation Biology as part of a special section on citizen science. Patience has never been my strong suit, so I’m overjoyed to announce that that special section is finally published!
The paper takes a pretty detailed look at how we turn your answers into our final dataset (that same one that was published in Nature Scientific Data last June). Remember that you guys are good, and even when you’re not sure about what you’re seeing, those wrong answers help us determine just how difficult an image is. My favourite demonstration of this is the boxplot below:
Now, I’ve written about this guy (and how to read boxplots) before. The gist of it is that we calculate a measure of disagreement across all of your answers for a given image. The disagreement score (also called evenness) ranges from 0 to 1, with 0 meaning that everyone agreed on what they saw and 1 meaning everyone said something different. You can see from the histogram on the right side of the plot that the vast majority of images were easy: everyone said the same thing! A good number of images are easy-ish, and a very small portion of the images are hard, with high disagreement scores.
When we compare images to experts, and look a the disagreement scores for images that were identified correctly or incorrectly, we see that images that were correct have generally lower disagreement scores (box on the left) than those that were incorrect (box on the right). That means we can use the disagreement score to predict whether images are probably right or wrong. If an image has a high disagreement score, it’s probably wrong or impossible, and we might want to have an expert review it before using it in an analysis.
For example. Across all images with disagreement scores 0-1, we know that 97% of images are correct. But say we want higher accuracy, so we set a threshold of images we accept and target for review. For example, 98.2% of images with a disagreement score of <0.75 are correct, so we could just accept all the ones with scores <0.75 and target all images >0.75 for review. Looking at the histogram to the right, that’s a pretty small percent of images needing a second look.
If 98.2% isn’t good enough, we can make that threshold stricter.
99.7% of images with disagreement scores <0.5 are correct, so we could set that as our threshold, and conduct an expert review of all images with scores above that. It’s still a relatively small number of images we need to look at.
Anyway. I know I’ve written about this before, but I think this really gets at the heart of why the Zooniverse/Snapshot Serengeti approach works for producing useable scientific data. And why your answers, even when you’re not confident in them, are so incredibly valuable. This approach means that ecologists and conservation biologists can engage volunteers like you on other camera trapping projects to tackle their own enormous camera trap datasets – enabling us to do bigger, broader research much faster.
As always, this wouldn’t be possible without your help. So thank you, again, for time and your clicks. And I can’t wait to see what you help us discover next!
There are definitely pros to early-morning fieldwork! Heading out to do camera trap experiments often required hitting the road before sun-up, and with the right composition of clouds, you often got to experience beautiful sunrises. Here, you can see the front of my LandRover as we’re about to tackle this swampy stretch of road!
My favorite part of Africa has to be the decorating.
One of the latest projects Craig Packer has been collaborating on involves trying to study cooperative behavior in lions by tempting these big cats hunt different “toys” – like this life-sized wooden buffalo:
One of the most hilarious disasters of our last field season came as a result of trying to lug all of these over-sized ungulates across South Africa. Apparently, simply ratcheting them on to the roof of your truck is only good until you start going fast enough for the wind to rip up and under them (i.e. anything over about 40 miles an hour). I have great pictures of Craig trudging across the highway to retrieve bits and pieces of giant warthogs, wildebeest, and other large wooden creatures whose sudden appearance flying off of our car must have completely baffled our fellow motorists.
One of the neat carnivores I got to work with in South Africa that we don’t experience much in Tanzania is the African wild (or “painted”) dog. These endangered carnivores live and hunt in highly social packs which, like wolves, are dominated by an alpha male and female. African wild dogs used to roam the Serengeti, but vanished in the park in the early 1990s due in part to diseases such as rabies and canine distemper contracted from domestic dogs.
During my first field season, I was fortunate enough to watch a pack of these animals being re-introduced into the park, and several more releases have taken place since. In total, over 60 wild dogs are now recolonizing Serengeti — we haven’t seen any in our camera trap areas yet, but there are rumors that they might be wandering through soon!
The best part about having a new season of photographs for me is the chance to “visit” Serengeti from the comfort of my own office. My research plans for 2016 don’t involve any trips back to Tanzania (mostly, I’ll be finishing up some experiments down in South Africa instead), so leaving Serengeti this last year was a very bittersweet experience. On the plus side, I did manage to grab a lift on one of the small bush planes that fly across the park, and the views were spectacular!
We promise that this one was worth the wait – hot-off-the-press Season 9 is one of our largest seasons yet, with over a million images (okay, okay, 300,000 some-odd “capture events”) to explore from our last year in the field! Head on over to www.snaphshotserengeti.org to join the search!
As we move through this new season, we’ll be updating regularly to share results that are emerging from our recent analyses, describe the brand new teaching tools we’re developing using this data to introduce young scientists to the world of research, and let you meet some of the branch new faces joining the Snapshot Serengeti team. Don’t forget to ask questions, post cool and unique pictures, and follow us on Facebook and through our Zooniverse discussion boards!
Thank you again for all of your effort in looking through these photos — this project wouldn’t be possible without you.
Thanks to you, we’ve already plowed through the entire “lost season” of Snapshot Serengeti. Granted, it was pretty small – only about 16,780 or so subjects, but you guys really knocked it out of the park.
So, first, let me say THANK YOU for your help!! It’s always such a thrill to announce a new season of data, and then watch that progress bar blast forward.
Second, speaking of progress bars, some of you have noticed that this progress bar went from zero to gone in, well, no time at all. Turns out we had a small bug in our code that meant we couldn’t see the progress bar advance. So finishing the lost season really took us by surprise too.
Third, if you feel like you’ve gone all excited by a new season to classify, only to have it ripped away, I promise that things will get better soon. We’ve now finished the major pre-processing for Seasons 9 and 10, and just need a couple weeks of minor pre-processing/porting to the Zooniverse. We’ll get them up online in early 2016 and you can classify to your heart’s content. Hopefully it will be something to cheer us all up as we return to work after the holidays. Seasons 9 and 10 are both really big, so there’s no risk you’ll miss out on them!
In the meanwhile, if you need an animal fix, Chicago Wildlife Watch could use your help!