Engagement with Nature During COVID in India

1 Overview

India went into strict lockdown on March 25th, 2020 due to COVID-19 restricting its population to their homes.  All non-essential transport, schools, industries, and hospitality services were suspended. While the poor have suffered–especially migrant workers unable to travel home–this has also been a time of respite for the relatively privileged.

It is hoped that the quieter streets and cleaner air may bring people closer to nature. An app among the growing ‘COVID technology mix’ is eBird (probably not as growing as rapidly as zoom, Netflix, and houseparty) , a digital birdwatching diary and community for nature enthusiasts. I have been analyzing data from eBird for my research at UBC.

There are currently 20,000 registered eBird users in India who have together logged close to 1 million trips since 2014 (www.ebird.org/India). During March 2020, as lockdown was imposed, 6028 users were active. In this blog post, I summarize data from these individuals to get an idea of the engagement with nature during these unprecedented times.

1.1 Birdwatching Over Time

This plot shows the trend in the number of birdwatching trips during COVID. Observations are grouped according to whether the birdwatcher was stationary or travelling. After lockdown, we can consider stationary trips as “balcony birding”.

Before lockdown, travelling trips are more common, as individuals prefer going to ‘hotspots’ with more species and may combine birdwatching with hiking or walking. A few days before lockdown, this relationship flips. Travelling trips plummet near-zero because of the restrictions and the number of stationary trips mirrors pre-lockdown travelling levels as balcony birding becomes the new norm.

1.2 Number of Users

This plot shows the number of daily eBird users during COVID-19. Prior to the lockdown, usage is twice as high on Sundays compared to weekdays when activity is low because users are at work. After lockdown, users begin balcony birding throughout the week while they are confined to their homes. The post-lockdown peak corresponds to a Friday, whereas pre-lockdown Friday levels were almost half as much. Similarly, user numbers on the post-lockdown Monday and Tuesday are substantially higher than the corresponding weekdays before lockdown. Overall, this shows that birders continue to birdwatch from their homes after lockdown, perhaps as an escape from the current stress.

1.3 Species Diversity

Several articles have shown improvements in environmental quality – especially pollution – after lockdown, due to the halt in economic activity. There has also been anecdotal evidence of biodiversity increases and wildlife proliferation in urban habitats.

Using eBird to study species diversity is problematic because of endogenous data collection. eBird users possess a range of abilities (observer bias), and also travel to locations where species diversity is already the highest (location bias), making it difficult to infer a representative measure of species richness. Because of the strict lockdown, the location bias is somewhat controlled for since users can only record species around their homes. The observer bias is also partially dealt with since incorrect records logged by new users will be flagged by eBird’s rigorous vetting system.

1.3.1 Graphical Evidence

This plot shows time trends in the total number of unique species (species richness) observed across all eBird users in March 2020. Species richness on travelling trips plummets towards zero leading up to lockdown as outdoor trips stop. In contrast, we do not see evidence that species richness on stationary trips increases post-lockdown. This could be for two reasons. First, this plot shows total species richness across India, which masks a lot of useful information at lower levels of disaggregation. Second, there are only 6 days of post-lockdown data, which may not be long enough for species to (re)appear.

In this figure, we make data more comparable by restricting to stationary trips and comparing district-level species richness on the Saturday before and after lockdown. The boxplot shows mean species richness per district (across all users) is higher after lockdown, in line with citizen reports of rare(r) species reappearing.

1.3.2 Event Study

For the final description of species diversity during COVID, we conduct a more rigorous analysis. The graph below shows results from an event study regression of species richness per trip on time dummies in a 6-day window around the lockdown. We include district fixed effects to control for baseline biodiversity as well as trip duration since users may take longer trips and therefore see more species after lockdown. Each point represents mean species richness relative to the day of lockdown. Starting two days after lockdown, species richness shows an increasing trend, however, it is statistically indistinguishable from zero.

Overall, the evidence is mixed regarding species richness. More data is needed to properly identify trends before and after lockdown. We will update this analysis as more eBird data is released.