Chicago Wildlife Watch Talk

Science Questions

  • dr.whyte by dr.whyte

    Please post here if you've got any questions about the science goals of the projects! You may be curious about the wildlife you see or what happens to images after you've tagged them. Anything really!

    Posted

  • Silene by Silene

    Does it matter that I am shown an image more than once and therefore classify it many times?

    Posted

  • DZM by DZM admin

    Hi @Silene , are you certain that it's the exact same image? Many images are taken from the same camera trap and thus can look very similar.

    Posted

  • Silene by Silene

    Hi DZM, yes I am certain that I have classified the same image several times. Patrticularily the back of a deer. I discount apparent duplicate images of empty images.

    Posted

  • DZM by DZM admin

    I'm going to ask mfidino@lpzoo.org ... do we have duplicate issues that you know of?

    If you see this image again, give it a unique tag (say, #duplicatetest1), then check for that tag when you see it again to make sure that it is in fact the exact same image. Thanks!!

    Posted

  • mason_UWI by mason_UWI scientist

    We've never had any issues with duplicate photos, though there are times when photos look INCREDIBLY similar. The only thing that I could potentially think would cause this on our end is us accidentally setting a camera trap to take 3 photos in quick succession instead of the normal 1 that we do, but an error like that is incredibly rare. The cameras take photos every 30 seconds so if there a deer stays in the same spot at that time you could end up with a number of similar photos.

    Posted

  • Silene by Silene

    Thanks for the information.

    Posted

  • buffalogroveP by buffalogroveP

    Thanks for this discussion! I have been noticing the same things as Silene. I'm thinking that sometimes the random presentation of pics sends me the same pic to classify more than once --- other times it is a pic that is very similar. I've seen the deer back and "Arthur" a lot too. It's too tedious to keep checking each pic to see if anyone made a comment about it -- but I do checkout some pics sometimes. Others, just get classified as nothing there -- I used to mark the cars one as cars -- but stopped that since they are in the distance -- and I did see a squirrel in that spot at least once 😉 Also, the camera by the Cal-Sag doesn't get much....

    Posted

  • DZM by DZM admin

    I love that "Arthur" is actually becoming a thing. Citizen science is awesome. 😄

    Posted

  • escholzia by escholzia

    I logged the IDs of 200 images from one camera - there were no repeats, although many of them appeared identical.On the other hand, I've seen one unique image (several athletic shoes running past a camera) at least five times. Several possibilities here: there are repeats, the camera triggers multiple shots, or someone was trolling us. My guess is that some cameras are just over-sensitive and are triggered multiple times by the same motion.

    Posted

  • escholzia by escholzia

    I'm curious about how the researchers count animals. Since the cameras are likely to pick up animals on their home turf far more often than those that are passing through, how do they interpret the counts? For example, I'd guess that 90% of the "mouse" counts are the one single mouse that hangs around under one set of shrubs.

    Posted

  • DZM by DZM admin

    A good question! Perhaps this information is more useful in determining an animal's range than it is for population?

    Would definitely be interested in a scientist's opinion here.

    Posted

  • mason_UWI by mason_UWI scientist

    Great question! It is rather difficult to infer abundance data from camera traps for the exact reason you explained: we are more likely to get repeated photos of individuals which will bias count data. For some species, like a zebra or cheetah, you can get around this issue because an individual has unique markings, whether they be stripes or dots. However, our urban wildlife community in Chicago does not contain species that we could individually identify. As such, we do not look to count data for our analysis and are more interested in site occupancy (whether or not we observed a particular species at a specific site). Both occupancy and abundance are related, as the probability of occupancy is simply just the probability that the abundance of a site is greater than zero, and we can still ask very interesting questions regarding patch occupancy instead of overall abundance in an area. For example, we can ask if the presence of one species impacts the likelihood of another being present at a site or if a sites proximity to a natural water source increases the likelihood of a species being present.

    I hope that answered your question, if you are interested in this topic I could write up a blog post on it.

    -Mason

    Posted

  • escholzia by escholzia

    Thanks - that answers my question.

    Posted

  • camelsamba by camelsamba

    I have a question about uncertain identifications. For example, I just got a photo where I know something is there because I see eyeshine, but that's the only detail I can pick out. It doesn't seem right to mark it as "nothing here" but i have no clue what animal it is... (i also don't know how to link to specific images, so i can't show which it is)

    Posted

  • mason_UWI by mason_UWI scientist

    In those cases, try to make an educated guess. Photos without a lot of agreement between users will get flagged for us to take a look at them. In this case, if even we cannot decide what is in the photo we will tag it as unknown after it has been classified on chicago wildlife watch.

    Things to ask yourself: how large is the eyeshine? Can you guess the size of the animal is from the eyeshine? I'm pretty good with deer, mouse, flying squirrel, and rabbit eyeshine, but the rest can get pretty difficult.

    It may actually be a good idea for us to put up a few examples of species specific eyeshine for everyone in a blog post. Thanks for asking about this!

    Posted

  • camelsamba by camelsamba

    yes, some examples of eyeshine would be helpful! i've heard from a mammalogist that there are some helpful rubrics, although i don't know how many of them would apply to photographs.

    Posted