Urban Biodiversity and Demographics

Urban Biodiversity and Human Wellbeing

In 2018 and 2019, nine cities and towns across the country participated in a pilot project using the Urban Biodiversity Information Framework (UBIF) to develop new ways to inventory their biodiversity using the tools of citizen science—smartphones and the iNaturalist app.

But the representations from each city in the pilot also wanted to ask questions beyond just how to measure urban biodiversity for its own sake: How do people value urban biodiversity and what benefits does urban biodiversity offer to people?

Asking how parks benefit people, and how and why people value parks, is not a new line of inquiry.

For at least a decade, park advocates across the country have built the case that people are happier and healthier when they have accessible parklands nearby. The National Park Service has built a ParkRx program to gather evidence and help communities build the case that outdoor recreation leads to improved health. And the Trust for Public Land has built the massive ParkServe database to score cities on their park accessibility and to highlight areas of critical park need. States like California have done extensive work to incorporate measurements of park need into the metrics used to distribute state funding.

The efforts to connect park access to human health — and then measure park access — rest fundamentally on access to demographic data (the Census) and comprehensive park inventories that aggregate park boundary and access data from dozens or even thousands of jurisdictions. This includes efforts like the Protected Areas Database of the United States, Colorado’s COMAP, and the California Protected Areas Database.

But such efforts say nothing about biodiversity. A pocket park with a bit of grass and a swingset counts as much to park access as an oak forest with a creek running through it. If nearby parks are mostly mowed ball fields and playgrounds, are the people in that neighborhood worse off than if there were parks with a range of more natural habitats nearby?

The answer to that question is probably “Yes,” but not for the reason urban nature advocates would like: Wealthier areas tend to have more diverse vegetation according to at least one recent study. The authors of that study write, “[T]he distribution of urban vegetation, and residents’ access to it, should be subjected to an equity analysis on a larger scale.”

With the UBIF project’s emphasis on using citizen science tools to survey biodiversity more rigorously over the next [X] years, we also saw an opportunity to measure who is doing that citizen science measuring and where they are doing it.

Citizen Science Data as Social Data

Anecdotally in Austin, St Louis, and Pittsburgh, we see evidence of an expected gap where the presence of iNaturalist data seems to indicate income at least as much as it does biodiversity. This is virtually the only common thread among the cities and towns in the pilot program.

By drawing on a few key Census variables, can we give cities tools to understand their citizen science user base? If only wealthier areas are getting more iNaturalist data and/or if the number of users overall is not growing in some parts of the city, what does that say about how urban biodiversity measurement efforts can be made more relevant and useful to a broader diversity of the people in each city?

Taking this approach here, we are not measuring biodiversity. We are measuring citizen science activity. As cities expand how they use iNaturalist, we’d expect it to become a true proxy for urban biodiversity, but for now, we are looking at iNaturalist user data as a social data set along the following lines:

  • Observation count per tract
  • Species/taxonomic unit count per tract (iNaturalist counts “taxa,” meaning a spot along the tree of life from kingdom down to subspecies)
  • Unique dates per tract, to measure whether use patterns seem to fall into concentrated events (fewer dates) or more regular use (more dates)
  • User count per tract overall and per year since 2008, to see if more people are contributing over time

And then from the Census, we are looking at the following:

  • Median household income
  • Percent in poverty
  • Percent by race and ethnicity

Each pilot city has a page on this website, where we show all the data and suggest some interesting patterns that emerge unique to each city. We also present a special per-city highlight dataset that represents some special area of interest or focal campaign related to urban biodiversity and community engagement in each place.

Data Glossary

VariableMethodology
Median Household IncomeFrom American Community Survey. ACS 2013-2017. Field B19013.
Percent in PovertyPoverty count divided by total population assessed for poverty.
Percent WhiteRace count divided by Total Population for the tract.
Percent AsianRace count divided by Total Population for the tract.
Percent African AmericanRace count divided by Total Population for the tract.
Percent WhiteRace count divided by Total Population for the tract.
Percent Hispanic/LatinoRace count divided by Total Population for the tract.
Percent Native Hawaiian and Pacific IslanderRace count divided by Total Population for the tract.
Percent American IndianRace count divided by Total Population for the tract.
Percent Other RaceRace count divided by Total Population for the tract.
Percent 2 Or More RacesRace count divided by Total Population for the tract.
Observation CountNumber of iNaturalist observations per tract. Filtered observations to those where obscured = False.
Taxa CountNumber of unique taxa per tract. Filtered observations to those where obscured = False, verifiable = True, and have taxon_id.
Unique Users In 2008Number of unique iNaturalist users per tract for 2008 only. Filtered observations to those where obscured = False.
Unique Users In 2009Number of unique iNaturalist users per tract for 2009 only. Filtered observations to those where obscured = False.
Unique Users In 2010Number of unique iNaturalist users per tract for 2010 only. Filtered observations to those where obscured = False.
Unique Users In 2011Number of unique iNaturalist users per tract for 2011 only. Filtered observations to those where obscured = False.
Unique Users In 2012Number of unique iNaturalist users per tract for 2012 only. Filtered observations to those where obscured = False.
Unique Users In 2013Number of unique iNaturalist users per tract for 2013 only. Filtered observations to those where obscured = False.
Unique Users In 2014Number of unique iNaturalist users per tract for 2014 only. Filtered observations to those where obscured = False.
Unique Users In 2015Number of unique iNaturalist users per tract for 2015 only. Filtered observations to those where obscured = False.
Unique Users In 2016Number of unique iNaturalist users per tract for 2016 only. Filtered observations to those where obscured = False.
Unique Users In 2017Number of unique iNaturalist users per tract for 2017 only. Filtered observations to those where obscured = False.
Unique Users In 2018Number of unique iNaturalist users per tract for 2018 only. Filtered observations to those where obscured = False.
Unique Users In All YearsNumber of unique iNaturalist users per tract for all years combined. Filtered observations to those where obscured = False.
Unique Dates In All YearsNumber of iNaturalist unique dates per tract for all years combined. Filtered observations to those where obscured = False and where date is not null.