Monday, August 27, 2018

What can we learn about chimpanzee groups from camera trap videos?

Hey Chimp&See enthusiasts! We recently published a paper about estimating chimpanzee group characteristics using camera traps, and I would like to take the opportunity to share our findings with you and give a little information about what these results mean for the work we do on Chimp&See. 
Our paper, entitled “An assessment of the efficacy of camera traps for studying demographic composition and variation in chimpanzees (Pan troglodytes)”, was recently published in the American Journal of Primatology.

Photo credit: MPI-EVA

Have you ever wished there were a way to know whether the camera traps actually record everything that’s going on when chimpanzees are present? Have you sometimes wondered whether more chimpanzees are just off camera nearby? Have you noticed how some individuals rarely seem to appear on camera, but others pop up in loads of different clips?

In this study, we were able to compare who appeared on camera to who was actually present, and to determine whether the camera traps were a good way to identify all chimpanzees and measure seasonal variation in party size. We looked at this by comparing two data sets from the same chimpanzees at Taї National Park, Côte d’Ivoire, recorded at the same time. One data set came from camera trap videos while the other came from direct observational data collected by researchers (which we know to be a good reflection of which chimpanzees were actually present and how big the parties actually were). By comparing the data we got from both sources, we found a few key things…

1)    We missed a lot of individuals in the camera trap videos!
If we considered the sum of all individuals recorded together in a given camera trap event, we missed about half of the party on average. So when we view clips on Chimp&See (especially individual 15-second clips), it’s good to always be aware that we’re often glimpsing through a small window into a bigger, more complex scene that’s unfolding out of view. This is a good reminder not to assume too much as we watch chimpanzee clips.

2)    Despite missing some information, we’re still getting really cool and useful data!
Chimpanzee party sizes tend to vary seasonally, based on factors like resource availability and the presence of estrous females. In this study, we found a similar pattern of seasonal variation in party size in both the camera trap and observational data, despite the fact that camera traps often missed some members of the parties. This highlights how useful the camera trap data can be despite incomplete information, and lets us know that the Chimp&See data also offer the potential to look at party size variation in all sorts of ways. So in case you’re wondering, all that chimpanzee counting you’ve done is still extremely useful!

3)    Those chimpanzee IDs are so valuable!
We also looked at whether all individuals appeared on camera eventually, and whether events like immigrations, emigrations, births, and deaths could be detected in the camera trap data. We found that all chimpanzees except a newborn infant born near the end of the study appeared on camera, and that deaths and an emigration were detected via disappearance from the camera trap data. We could accurately figure out the chimpanzee community’s composition based on the camera trap data, too. This information all relied on the chimpanzee IDs assigned in the videos (in this case, by researchers who were directly familiar with the chimpanzees, having studied them). On Chimp&See, the only way to obtain this very valuable information is via the agreed-upon chimpanzee IDs that citizen scientists like you help provide. This information is extremely valuable to the project, and the only way we can answer some critical questions about the chimpanzees you see on video—a HUGE thanks to you! 

Overall, this study gives us some insight into how valuable camera trap data are for understanding chimpanzee communities, as well as shedding light on the ways in which we should be careful when considering what we can get from the videos. We’re so grateful for all the ways you’ve contributed to Chimp&See and can’t wait to share more results with you in the future! Keep up the great work!

No comments:

Post a Comment