A camera trap grid runs through it: surveying Etosha wildlife large scale and long term!
Updated: Feb 7
One of the many challenges that wildlife biologists and managers face is to know how many animals occur in a given area, how they distribute themselves in space and how this distribution changes with seasons. Estimating ungulate population size and structure and their spatial distribution is traditionally done using road or aerial surveys. However, when one is also interested in how these 3 parameters vary in time, at the seasonal or annual scales for instance, these survey methods quickly become extremely costly and logistically unpractical, especially over an area as large as the entire Etosha National Park (~23 000 km2)! Therefore, we decided to use another method, and deployed a total of 83 camera traps on a 10x10km grid to cover the Park. These cameras are deployed on trees, facing a wildlife trail to maximise capture rate. Of course, these cameras take pictures of anything moving in front of them, so we are not only surveying ungulates, but any other species moving in the Park. They each record 3 images once they are triggered in the hope of having at least one clear shot allowing for individual identification of species such as leopard, cheetah and giraffe.
The first part of the grid (in the central area) has been deployed and maintained by our colleague Wendy Turner’s team (USGS) since 2018. ORC has now teamed up with Wendy and Jim Beasley (University of Georgia) to extend this grid over the entire National Park. We deployed the Eastern grid in September 2022 and the Western Grid in November 2022 and January 2023. Each camera needs to be serviced (i.e. replacement of memory card and batteries) about every 2 months. That is a lot of work for a lot of images, but also a lot of time spent in the park, which is always a bous!
Back in the office, images are backed up on a hard drive and later uploaded on the Trap Tagger platform (https://traptagger.co.uk) to be processed with the help of artificial intelligence (AI). This is a great time saver as the AI algorithms are very fast and reliable for detecting empty frames (e.g. camera triggered by a branch swaying in the wind) which allows us to view only images in which an animal was actually detected.
Using the data, our goal is to obtain a much clearer idea of ungulate numbers as well as their population structure, but most importantly of their seasonal movements in response to the rainfall at a large spatial scale. We also aim at determining the occurrence and distribution of smaller and seldom seen species such as Bat eared fox, African wild cat, Aardwolf and Honey badger. This data will also be used to produce population size estimate for leopard and cheetah for instance. But we won’t stop at this, and the data generated by such an extensive survey opens the door to answering many more questions!
Stay tuned for more updates as we start to work our way through the images!
A glimpse through the traps...
Snapshots from the field