Recapping the 2017 City Nature Challenge

This post is brought to you by Marschall Furman, a first year PhD student in statistics at NCSU.  Marschall is a huge sports fan and proud owner of 3 baby cacti. Thanks, Marschall!

From April 14-18 this year, thousands of citizen scientists across the US took part in the 2017 City Nature Challenge on iNaturalist. Volunteers from 16 different cities contributed by taking pictures of wildlife in their local environments. In total, there were more than 125,000 photos added to this project in only 5 days.

The competition was strongly dominated by the larger cities in California and Texas, followed by a steep drop off. We can break these totals down at the participant level from each city to compare how much effort they each put in.

There were more than 4,000 contributors in total, and the overall median number of photos taken was 5. We also see that many of the more involved cities had larger participation cohorts, and some even had “super-volunteers” that uploaded over 1,000 photos.

One potential use for iNaturalist projects like this one is an improved understanding of species diversity and richness. In order for these photos to be deemed sufficient for scientific research, the species identified in the photo must be agreed upon by at least one other user. In the dataset, each photo has a designated quality grade (e.g. “Research Grade”) that indicates its validity. In the next graph, we take a look at the quality grades for each city that participated in the challenge.

The “Casual” label indicates that the species in the photos are unable to be validated, while “Needs ID” represents photos that are verifiable, but have not yet been examined by other users. A positive takeaway is that most of the cities have small percentages of “Casual” photos (< 20%), so most of them could be useful once properly identified. We also see that some cities with lower participation also have higher proportions of “Needs ID” photos, suggesting that they don’t have as large of a group to help verify findings.

When participants upload a photo to iNaturalist, they are also asked to include identification information about the species they snapped. This information allows us to investigate the dataset as a whole, as well as the diversity of any particular city. In total, 90 different taxonomic classes were spotted. Below are the top classes reported, both overall and in the Triangle area.

We can see that most of the photos uploaded were taken of Magnoliopsida, flowering plants, and Insecta, insects. There are probably a couple of reasons for this.  First, both groups are abundant and either don’t move or are known to move slowly, which makes them easier to take photos of.  Additionally, they can be quite interesting, attractive, and identifiable species.

We see some distinct differences between the NC data and the larger project, such as the lower prevalence of Aves (birds) – 5.8% in NC vs 10.5% nationally – and the increased presence of other classes like Filicopsida (ferns). Also of note is that roughly 12% of the photos contributed are lacking this identification, and further investigation of these photos may be required.

We can start to understand the species richness of the participating cities by looking at how many species each of them identified. Across all participating cities, there were roughly 8,000 different species identified. Here are the most common occurrences for the whole project.

Interestingly, although the flowering plants were the most commonly identified class, they only had 2 of the top 10 most sighted species. This suggests that there was a wider breadth of plant photos contributed; in fact, there were almost twice as many species of Magnoliopsida found (~3400) as those in Insecta (~1800), the next closest class. Lastly, we can check out the ten most identified species in a few of the most prevalent classes.

Many of these species are ones we encounter on a regular basis or can be easily identified by from their appearance, so it makes sense that they are the most common sightings of wildlife across the country. Which ones have you spotted in your environment?

Thank you to all of the participants from the Triangle area and around the US.  We hope to see you back next year for City Nature Challenge 2018!

Note that participant data is accurate as of April 23, 2017, taxonomic data is accurate as of May 25, 2017, and the “NC Triangle” label refers to the whole Raleigh-Durham area.