Introduction
Emergence of the COVID-19 global pandemic severely altered the lives of people worldwide (). Governments responded by declaring and implementing restrictions on social life aimed at curbing the spread of the virus, including mandatory lockdowns, the closure of public spaces, and social distancing (). The combination of COVID-19 and associated restrictions resulted in widespread mental health issues such as depression, anxiety, stress, boredom, and social isolation (; ; ).
COVID-19 control measures reduced access to real-world social infrastructure. Pandemic restrictions limited social activity and strained community cohesion (). It closed schools, brought travel to a stand-still (), and shifted activities to virtual environments, increasing the importance of the internet as social infrastructure (). Enforced physical isolation catalyzed new approaches and infrastructure (such as the widespread adoption of Zoom among the general public) that could boost social connectedness.
Families and individuals sought out alternative ways to connect with the outside world. This included observing nature () as well as increasing engagement in online communities—both of which can be provided by participation in citizen and community science (CCS) (). For example, CCS project introduction events, conceived as in-person events at public libraries, moved to Zoom and were open to anyone who registered. SciStarter, with support from the National Library of Medicine, coordinated close to 100 of these Zoom events for Citizen Science Month in April 2020 during the height of lockdowns.
There are many terms for public participation in research including citizen science (CS) and community science (; ). These terms carry different meanings and importance for different audiences. We use the term CCS to include a range of approaches. We acknowledge the importance of the term community science as emanating from social justice–oriented projects () to address issues facing vulnerable communities. Citizen science, as the term has been used since the mid-1990s, focuses on institution- or researcher-led projects harnessing the data collection capacity of volunteers (), and/or projects created by educational institutions with a primary purpose of engaging communities in science. For this paper, we analyzed projects that fit the above definition of CS, though our findings may apply to CCS more broadly.
Participation in CCS has been shown to offer numerous mental and emotional benefits () and increased sense of belonging to a community (). Creating opportunities for people to be involved in primary data collection can empower them and improve the body of knowledge by engaging a variety of backgrounds and viewpoints (). This empowerment improves science literacy, confidence, and trust (). Nature-based projects can improve wellbeing through increasing participant connectedness to nature (; ).
Increased community engagement in the study and management of natural resources may also increase community resilience (; ). Resilience has a variety of definitions and interpretations () that relate to the concept of adapting to change and continuing to exist under new conditions (). Strategies incorporate and integrate resilience among linked social–ecological–technological systems (SETS) (). By providing more complete data, especially from hard-to-access areas such as urban backyards (), CCS can improve conservation and restoration efforts () that improve ecological resilience (). CCS can support social resilience by increasing the knowledge and competency of community members who make decisions that affect SETS. Co-discovery, by scientists, managers, and citizens, of the processes that cultivate resilience is an important element of sustainability (). Understanding how projects adapted to the unique situation brought about by COVID-19 provides insights as to how CCS can increase community resilience.
Interest and participation in CS boomed during the initial months of the pandemic. Contributions to popular platforms such as eBird, iNaturalist (), and Zooniverse () increased. On SciStarter (an online CS hub containing over 2,000 projects), there was a 480% increase in participation in April 2020 compared with April 2019 (). The reasons behind this are myriad and interconnected, including desire for intellectual stimulation, need for social connection, and concern about the environment (). The transition to online distance learning motivated parents to search for science education opportunities. Scientists turned to local volunteers to accomplish data collection that was no longer easily done by paid students and technicians because of social distancing requirements ().
This trend of increased participation was not universal and varied with the type of project. Crimmins et al. () found that some large-scale biodiversity monitoring projects witnessed increased participation (iNaturalist, eBird), whereas participation in other such projects decreased (eButterly, Nature’s Notebook). Basile et al. () found no major change in the number of observations submitted to iNaturalist by birdwatchers in Italy and Spain. Large group projects were particularly impeded by social distancing mandates and witnessed a decrease in the number of new participants (). Kishimoto and Kobori () found a 63% decrease in participants in the City Nature Challenge Tokyo in 2020 compared with 2019, though they saw an increase in identifications—an activity that can be done by individuals online.
The severity of lockdowns imposed by governments also resulted in differences in participation between different locations (; ). Even for projects in which individuals continued to participate, the location changed, with more people participating in projects that could be done at or near home rather than traveling to distant places (; ; ). Amongst these analyses, the question remains as to how participation in different types of CS projects was affected by COVID-19 restrictions.
To explore this question further, we had two objectives. Our first was to document the impact of the COVID-19 pandemic on participation in existing CS projects classified along a spectrum from in-home to public natural areas. Our second was to explore how participation in CS projects varied with geographically different government restrictions.
We hypothesized that differences in where projects occurred and how the data needed to be collected would determine whether there was an increase or a decrease in participation. Specifically, projects that could be conducted at home, whether online only or through making observations in or near home, would experience increased participation, while those that required a coordinated event or access to public spaces would see a decrease. We further hypothesized that there would be more participation in projects that could occur away from home in United States (US) states with fewer COVID-19 restrictions and more participation in at-home projects in states with many COVID-19 restrictions.
Methods
We employed a mixed-methods approach. This consisted of examining digital trace data on CS participation from SciStarter.org before and during the pandemic, constructing a quantitative index to measure how restricted daily life in each US state was during COVID-19, and semi-structured conversations with project leaders to identify ways individual projects adapted to COVID-19 lifestyles.
Project data
We examined the activity of SciStarter users. SciStarter is comprehensive hub for citizen science where potential volunteers can search for projects of interest and track their participation. The SciStarter database includes more than 2,000 projects and has close to 100,000 registered community members. We included projects that received at least one data contribution between March 19 and May 31, 2020. This time period began with the date the first statewide stay-at-home order was initiated in the US and ended when most states had loosened some of the initial COVID-19 restrictions and daily case numbers were relatively stable.
There were 93 unique CS projects that received at least one data contribution during this time period. We coded projects into mutually exclusive groups based on characteristics hypothesized to affect participation trends (Figure 1). The four groups of projects were: “online,” for which the raw data had been previously collected and participants could observe, classify, and record data exclusively online (n = 27); “away from home” that required the participant to collect data in a public location, such as a park or coastline (n = 20); “at home indoor” that required the individual to collect data from inside their house (n = 13); and “at home outdoor,” where participants were required to collect data outside in their yard or nearby neighborhood (n = 33).
We classified citizen projects on SciStarter into four distinct groups according to this classification scheme. Rectangles indicate the primary questions used to classify projects at each level. Ovals indicate examples of projects in each classification group. Group names are provided in quotes below ovals.
Our data set was limited to projects in the US. SciStarter has been most actively used by projects in the US, and focusing on a single country allowed us to limit the COVID-19 policies we assessed. The results of our study directly apply to project design and participation in the US, although insights may be relevant to similar countries.
Participant data
We obtained activity data from SciStarter on participants who made at least one contribution during this time period (n = 3,602). The activity data consisted of the number of project “joins” (i.e., how many new projects an individual joined) and the number of project “contributions” (i.e., how many data contributions an individual submitted to a project). We created a third variable for “engagement,” which was a binary variable scored as a “1” if the individual made > 0 contributions to the project. This was created because effort required per contribution varies between projects, and this variable captured which projects participants were engaged in in some form. “Joins” thus represented new interest in projects, whereas “engagement” represented total activity for a project. We then combined the participant data with the project data to calculate the number of participants who joined and engaged with each project during our study period.
Additionally, we established a baseline by obtaining activity data from all participants who joined at least one project during January 1 through March 18, 2020. This represented trends in project joins prior to the start of pandemic, allowing comparison with data collected during the pandemic. A total of 1,955 participants joined 101 unique projects during this time period. We refer to joins from January 1 through March 18, 2020 as “pre-COVID-19” data, and joins from March 19 through May 31, 2020 as “COVID-19” data. We chose to use data on joins from earlier in 2020 rather than the same time period in an earlier year owing to the explosive growth of projects on SciStarter from 2019 to 2020. March 19 through May 31, 2019 saw only 323 project joins across 15 different projects. Thus, the data from January 1 through March 18, 2020 more closely resembles the COVID-19 data and serves as a better baseline.
COVID-19 restriction data
US states enacted varying restrictions on daily activities in response to COVID-19. To examine how this impacted CS participation, we categorized data on COVID-19 restrictions from the following sources: Kaiser Family Foundation (); National Academy for State Health and Policy (), and The Council of State Governments (). We used several restrictions as variables: whether there was a state-wide mandate requiring facial coverings; how many executive orders were issued; whether a stay-at-home order was enacted; the date a state first began to lift some initial restrictions on businesses openings; and whether a state had travel restrictions against out-of-state visitors. While we did not hypothesize these variables would necessarily explain variation in CS participation, they were used to gauge how restricted life in each state was.
We created an additive index for the total number of restrictions per state by coding our selected variables into binary categories. Mask mandate was coded as “1” if present statewide. Executive orders was coded as “1” if a state had issued more than 40, the mean number of executive orders issued over our study period. Stay-at-home order was coded as “1” if there was a state-wide stay-at-home order in effect. The date of first re-opening was coded as a “1” if states did not begin easing restrictions on businesses before May 8 (the mean re-opening date for states). Travel restrictions was coded as a “1” if travel restrictions for visitors were in place. This produced 6 possible groups a state could belong to (Table 1).
Table 1
Grouping of states according to the total number of COVID-19 related restrictions during March 19 through May 31, 2020. Possible restrictions included whether a state-wide mask mandate was required, the number of executive orders issued by the state, whether a stay-at-home order was enacted, the date a state first began to lift some of the initial restrictions on which businesses can be open, and whether a state had travel restrictions in place for visitors arriving from outside the state.
| NUMBER OF RESTRICTIONS | STATES |
|---|---|
| 0 | IA, SD, UT, WY |
| 1 | AL, IN, MS, MO, NE |
| 2 | AK, AR, CO, FL, GA, ID, LA, MN, MT, NV, NC, ND, OK, OR, SC, TN, TX, VT, WI |
| 3 | AZ, CA, HI, KS, KY, MI, NH, NM, OH, PA, WA, WV |
| 4 | DE, IL, ME, VA |
| 5 | CT, MD, MA, NJ, NY, RI |
We further reclassified states into two classes based on the total number of restrictions. States with 5, 4, or 3 restrictions were grouped together as “highly restricted states” (n = 22 states), whereas states with 2, 1, or 0 restrictions were grouped together as “minimally restricted states” (n = 28 states). This allowed us to have two classes with similar numbers of members.
Because participants are not required to specify their location when registering with SciStarter, our sample was restricted to those who voluntarily supplied location in their SciStarter profile (n = 232). They represented 44 US states. We obtained location data for 129 individuals from highly restricted states and 103 individuals from minimally restricted states.
Analytic methods
We used both project joins and project engagements as dependent variables to assess how the pandemic affected participation in different types of CS projects. Both dependent variables were log transformed to correct for a heavy right-skew in the distribution. We utilized R () to perform Kruskal-Wallis tests to examine differences in mean number of joins and engagements among the four project groups. We examined differences between the groups using pairwise Wilcox tests with the Bonferroni correction to adjust for the number of comparisons. We used t-tests to explore differences in mean number of project joins among the different project types between the pre-COVID-19 and COVID-19 data. We used Kruskal-Wallis tests, chi-squared tests, and t-tests in R to explore how the number of joins and engagements among the four project groups differed between highly restricted states and minimally restricted states. We considered significant differences to be P ≤ 0.05 for all analyses. We report means and standard deviations (sd) when appropriate.
Illustrative interviews
To provide depth and context to our quantitative data, we selected five projects that exemplified the different CS project groups and conducted ~1-hour semi-structured interviews with project leads. We selected projects from the SciStarter database that had the highest (or one of the highest) number of joins in each project category. These included Phylo (“online,” 144 joins), Stream Selfie (“away from home,” 183 joins), and Crowd the Tap (“at home indoor,” 132 joins). We also selected an extreme example of an “away from home” project not in the SciStarter database, Grunion Greeters, which requires observing fish behavior under very specific conditions. We selected ecoEXPLORE, a plant and wildlife observation project aimed at youth, as an example of “at home outdoor,” although it could also be conducted away from home. One researcher conducted all interviews via Zoom in April and May 2021. For Crowd the Tap, a member of our author team serves as the project director and provided information here according to the interview script. Interview questions focused on whether project leaders saw changes in participation, and whether they modified aspects of the project (training, data collection methods, etc.) in response to COVID-19 restrictions. Analysis consisted of comparison between project leaders’ perceptions of participation and our quantitative findings, and identifying similarities and differences in the ways projects were or were not modified.
Results
Impacts of the pandemic on participation in existing citizen science projects
SciStarter members joined a total of 68 unique projects over the study period (22 “online,” 12 “away from home,” 10 “at home indoor,” and 24 “at home outdoor”). Kruskal-Wallis tests revealed a significant difference in the mean number of joins per project among the four different groups (𝛸2 = 8.418, p = 0.038; Figure 2). “Away from home” projects had the lowest mean number of joins (1.15, sd = 1.63), while “at home indoor” projects had the highest mean number of joins (2.54, sd = 1.45). No significantly different pairwise comparisons between these group means were found (Figure 2). However, the pairwise comparisons between “away from home” and “at home indoor” (p = 0.091) and between “away from home” and “at home outdoor’’ (p = 0.056) were significant at a P ≤ 0.10 level.
We found a significant difference in the mean number of joins (log scale) per project among the four project groups during the COVID-19 pandemic (𝛸2 = 8,418, p = 0.038) but no statistically significant pairwise differences. Boxes represent the interquartile range (25th percentile–75th percentile). Horizontal bar in each box represents the median. Vertical bars extend downward and upward from boxes to the minimum and maximum non-outlier values, respectively. Individual points represent outlier values.
Number of engagements was highly correlated with number of joins (r = 0.93). Kruskal-Wallis tests revealed there was also a significant difference in the mean number of engagements per project among the four different project groups (𝛸2 = 8.042, p = 0.045; Figure 3), with “away from home” projects having the lowest mean number of engagements (0.529, sd = 3.53) and “at home indoor” projects having the highest mean number of engagements (1.500, sd = 2.98). No pairwise comparisons were significantly different (Figure 3). The pairwise comparisons between “online” and “away from home” (p = 0.061) and between “away from home” and “at home indoor” (p = 0.080) were significant at a P ≤ 0.10 level.
We found a significant difference in the mean number of engagements per project (log scale) among the four project groups during the COVID-19 pandemic (𝛸2 = 8.042, p = 0.045) but no statistically significant pairwise differences. Boxes represent the interquartile range (25th percentile–75th percentile). Horizontal bar in each box represents the median. Vertical bars extend downward and upward from boxes to the minimum and maximum non-outlier values, respectively. Individual points represent outlier values.
In contrast to the COVID-19 data, Kruskal-Wallis tests revealed no significant difference in the mean number of joins per project among the four different project groups in the pre-COVID-19 data (𝛸2 = 5.397, p = 0.145). T-tests revealed no significant differences in mean number of joins per project type between the pre-COVID-19 and COVID-19 data (Figure 4). While not statistically significant, there was a visually notable decrease in mean number of joins per project for “away from home” projects in the COVID-19 data compared with the pre-COVID-19 data (p = 0.201).
There were no significant differences in the mean number of joins (log scale) per project among the four project groups between the pre-COVID-19 and COVID-19 data. Boxes represent the interquartile range (25th percentile–75th percentile). Horizontal bar in each box represents the median. Vertical bars extend downward and upward from boxes to the minimum and maximum non-outlier values, respectively. Individual points represent outlier values.
Chi-squared tests revealed no significant differences in the total number of engagements for different project groups between minimally restricted and highly restricted states (𝛸2 = 6.267, p = 0.201; Figure 5). T-tests revealed that there were no differences in total number of joins by project group between minimally and highly restricted states for “online” (p = 0.672), “away from home” (p = 0.168), “at home indoor” (p = 0.788), and “at home outdoor” (p = 0.207). The mean number of total engagements per state was slightly higher in highly restricted states (6.79, sd = 6.65) than minimally restricted states (4.48, sd = 3.03), but Kruskal-Wallis tests revealed this difference was not statistically significant (𝛸2 = 0.489, p = 0.484).
There were no significant differences in the total number of engagements for different project groups between participants from minimally and highly restricted states during the COVID-19 pandemic (𝛸2 = 6.267, p = 0.201).
How project leads adapted projects in response to COVID-19
Leads for examples of each project type shared their impressions of how the pandemic affected participation. They also described any modifications they made in response, and any opportunities it created.
Entirely online project: Phylo
Phylo is an example of a CS project that takes place entirely online. Participants align acquired DNA sequences to help determine the evolutionary history of various species. There is a low barrier to entry into the project; other than access to the internet, a device that can support the program, and enough bandwidth to use it, project participants can participate anywhere, with no training. The project lead did not see an increase in participation under COVID-19 restrictions, but noted that there is always seasonal fluctuation, with more engagement during winter months or when there is press coverage of the project. No modifications were made or needed for the project to be available to participants during COVID-19-related restrictions.
At home indoor project: Crowd the Tap
Crowd the Tap represents a “at home indoor” project that is “one-and-done,” with a focus is on safe drinking water. Crowd the Tap participants answer questions about their pipes, water characteristics, and household characteristics to estimate risk of lead in their water. The project also collects water-chemistry data from those who request disposable chemistry strips. In the months before the pandemic, Crowd the Tap began preparing to work with two facilitator groups: librarians and teachers. The National Library of Medicine selected Crowd the Tap materials to be part of citizen science kits available in public libraries. The project director contracted with educational consultants for the creation of lesson plans for high school teachers. Initially, there was interest (e.g., inquiries) from librarians and teachers in the Crowd the Tap kits and lessons, but the PI suspected the pandemic negatively affected the ability of facilitator groups to follow through. More than 500 libraries ordered and received kits with Crowd the Tap materials and disposable chemistry strips. However, all libraries closed and reduced programming because of the pandemic. Only a few switched to online programs that featured Crowd the Tap, and only a few teachers requested disposable water chemistry strips for their students.
Away from home project with broad geographic inclusion: Stream Selfie
Stream Selfie, run by the Izaak Walton League of America (IWLA), is an example of a “away from home’’ project. Stream Selfie is a simple CS project designed to initiate public engagement in aquatic monitoring. The project can be done by an individual alone but does require collection of data at a stream location of their choice. The project leader saw an initial uptick in participation as COVID-19 restrictions were enacted, peaking through Citizen Science Month and Earth Day, with a subsequent drop to a level still higher than pre-pandemic. While the data collection procedure and interface did not change in response to COVID-19 restrictions, more detailed instructions were added to promote individuals joining through SciStarter. Recruitment messaging emphasized that the project was an outdoor activity amenable to social distancing. During the pandemic, the project leader put more effort on Stream Selfie compared with their other CS activities that required working in a group. Outreach also changed to encourage participants to investigate locations on their own property or at local parks rather than state and federal lands that were closed. Prior to the pandemic, most participants came from IWLA chapters, but since the pandemic and changes in project messaging, the project leader saw an increase in participants not associated with chapters. Participants noted how important it was to use the project to connect with nature during the pandemic.
Away from home project targeting specific locations: Grunion Greeters
Grunion Greeters, run by Pepperdine University and the Beach Ecology Coalition, represents an extreme example of a “away from home” project that requires access to specific sites at specific times in order to participate. California grunion, Leuresthes tenuis, are small marine fish who mate and deposit eggs on beaches from Tomales Bay, California to Ensenada, Baja California, Mexico from March to August during the highest high tides. These occur the night after the full and new moons, influenced by changes in the moon’s gravitational pull on the Earth’s oceans. Fish come up, or “run,” on the beach to deposit eggs that develop buried in the sand until the next high high tide when larvae emerge. Determined by this astronomic event, grunion run on the same nights near the same times across their geographic range. Volunteers are a vital component of the monitoring program, which would be technologically challenging and prohibitively expensive without CS. Participants collect data about the fish and conditions that could disturb them, such as light levels and late-night beach recreation.
In spring of 2020, California closed public beaches in response to COVID-19, coinciding with grunion spawning and rendering CS monitoring at most areas illegal. This greatly reduced volunteer participation early in the spawning season. The California Department of Fish and Wildlife removed the public schedule from their website and the Grunion Greeter project was not promoted, though participants from previous years engaged with project leads via email. Some volunteers were able to access the beach legally, such as staff from beach management agencies, lifeguards, beachside residents, etc. In fact, far more members of the public were observed on beaches at night than during previous years, despite the closures, which were not well enforced. The 2020 dataset was therefore small.
In spring of 2021, the project re-focused recruitment efforts online, and provided online training rather than live workshops. The data collection season was in progress when the interview was conducted and at the time of writing, but the project leader estimated that participation in 2021 will more than double that of 2020.
“At home outdoor” project focused on youth: ecoEXPLORE
Unlike the projects described above, ecoEXPLORE, run by the North Carolina Arboretum, is specifically aimed at youth ages 5 to 13. It is an example of an “at home outdoor” project, although it could also occur “away from home.” The project invites participants to observe flora and fauna, sometimes in response to specific badge-earning challenges to focus on a place (a hotspot) or taxon. The primary focus is engagement and education, but data are added to iNaturalist where they can be utilized for relevant research projects. Project leaders reported a rapid increase in participation and web traffic when the pandemic hit, which they attributed to school closures and the need for parents to oversee distance learning, though they also saw increased participation from adults who were not parents. Project leaders encouraged participation during the pandemic by shifting to web-based events and online training. This shift was possible because in-person educational activities at the Arboretum were cancelled, leaving staff more time to focus on ecoEXPLORE.
Prior to 2020, activities on ecoEXPLORE encouraged families to visit public parks, but the project was modified to encourage participants to observe nature within their local neighborhood. The move to encouraging independent local exploration also allowed the project to expand in geographic scope:
“We had already launched two regional hubs across the state prior to COVID. These organizations were already facilitating ecoEXPLORE within their local communities. I would say that COVID did allow for greater virtual participation, and it helped us reach places in the state that we were struggling to reach.”
Discussion
We found significant differences in participation based on key characteristics of the CS projects. Projects that required data collection at specific sites away from home saw fewer joins and engagements by SciStarter members than projects that could occur at or near home. This matched the pattern found by Basile et al. () of participants continuing to contribute data to projects but doing so from their neighborhoods rather than public locations. It was somewhat surprising that there were not more joins and engagements to “online” projects. This could be due to virtual meeting fatigue among people working from home who were searching for activities that took them away from a screen ().
An important limitation to the quantitative data is that we could not compare changes in joins and participation across years. Only 15 of the 68 projects in our sample were on SciStarter in 2019, impeding quantitative comparisons. We did compare joins in 2020 before and during COVID-19 to document a decrease in the number of joins for “away from home” projects. Our results still elucidate which types of projects received the most interest and engagement during the start of the COVID-19 pandemic.
The results of our quantitative analysis were well aligned with the experiences of project leads. Our exemplar of an “online project,” Phylo, did not experience the increase in participation we expected, which paralleled our quantitative findings. This may be because it was not heavily promoted during the restricted period as project leads were more actively promoting newer projects. Crowd the Tap, our example of an “at home indoor” project, also did not see the increase in participation project leads had hoped for based on pre-COVID-19 promotion. This project was designed to be facilitated by teachers utilizing hands-on materials, and was not easily re-designed for home use. Many affiliated teachers were overwhelmed and opted not to add ambitious assignments like CS. The lack of increasing participation may also reflect a preference among CS contributors to spend time outdoors.
Projects that occurred outdoors received the most joins and engagement. Participation in Stream Selfie increased as expected based on our initial hypothesis. This ran counter to the quantitative data. Stream Selfie was classified as “away from home,” as it required participants to search for a stream. However, further inspection of “away from home” projects revealed that Stream Selfie was an outlier with far more joins than other “away from home” projects. Stream Selfie differed from other projects in the “away from home’’ category. At the start of the pandemic, project managers encouraged more local exploration rather than visits to state or national lands. Project leads were able to capitalize on the additional interest due to their own increased availability. They further increased participation opportunities by marketing the project beyond IWL’s usual community. The flexible approach adopted by Stream Selfie contributed to its success during the start of the pandemic and suggests strategies that other “away from home” projects could employ to avoid decreased participation.
As an example of an “at home outdoor” project, ecoEXPLORE exemplified our quantitative findings, as it saw a large increase in participation, including an increase in geographic scale. Project leads were able to capitalize on a new need for science- and nature-based activities parents could do with their children in lieu of in-class experiences. In the cases of ecoEXPLORE and Stream Selfie, the pandemic not only influenced participation but also the nature of the projects themselves.
In contrast to the above projects, for which location could be easily adapted, Grunion Greeters was an extreme example of a project impossible for a majority of people to contribute to from home. This project, driven less by a desire for public engagement and more by a need for efficient and reliable data collection, was negatively impacted by access restrictions. In this case, CS was an important data collection methodology contributing to conservation and was dramatically hindered by the pandemic.
Despite the widely different restriction levels states imposed throughout the pandemic, we found no evidence for differences in CS participation. States with a higher level of restriction had slightly more engagement, though that result was not significant and could have correlated with many factors. We should note our analyses were limited by the fact that we only had location data for 232 of the 3,602 participants in our sample. Accessing location data for more participants could have produced different results.
Our research faced another confounding factor. Since 2016, a date in April on or around Earth Day has also been recognized as Citizen Science Day, an event created to increase awareness, participation, and benefits from CS activity across the US. In 2020, with support from the National Library of Medicine, that event expanded to Citizen Science Month, planned long before COVID-19 appeared. In the 11 years since SciStarter.org emerged and the field of CS formalized, participation in these events has grown steadily. It is difficult to tease apart how much of the observed increases in participation were due to pandemic-associated factors versus successful outreach.
Conclusion
It is well documented that CCS can be used to understand and respond to global change. CCS can be used to monitor extreme events such as flooding () and fire (), to examine impacts to biodiversity (; ), and to understand shifts in ecology and phenology (). It can be a powerful tool to detect and monitor emerging pests and invasive species () as well as infectious diseases (). Engaging in these projects can build individual confidence, capacity for stewardship, and social capital for conservation (). As more people rely on CCS for these important benefits, project leaders and the field as a whole need to be more agile and adaptable in the face of increasing uncertainty and disasters such as the COVID-19 pandemic. Project managers can adjust their approaches to sustain participation and to continue to provide these important community benefits. Our quantitative data demonstrates the shifts in participation that occurred, while the interviews suggest ways project managers adapted.
CCS can improve online communication infrastructure in a positive way that can increase community strength and resilience (). Community resilience to disaster is needed, particularly as climate-driven extreme weather and health events increase (). CCS can also contribute to community resilience through the utilization of the internet as social infrastructure () to foster connectedness during periods of physical isolation.
The year 2020 reintroduced us to the fact that global pandemics have massive societal impacts. CCS can be an important tool for building community and a positive connection between nature and society that, in the case of this pandemic, took on increased resonance. A global disturbance such as COVID-19 is a rare occurrence. However, it is likely that the world will need to face global challenges such as combating emerging diseases, responding to resource limitations, and addressing impacts of more frequent and severe weather events in the coming years. CCS’s ability to build connections may play an increasingly important role.
Intentionally building flexibility into CCS project design can create a more robust environment for participatory research. In order to ensure citizen science can be leveraged, we need to think about the characteristics of projects and how participation might be affected by major societal changes. Removing barriers, such as the need to attend in-person training, or the need for access to specific public spaces that may be difficult to reach, can increase inclusivity, broaden participation, and allow projects to continue to function under a range of disturbances. This goes hand-in-hand with efforts to expand broadband access. Though not all projects may be able to adopt this approach given their scientific objectives, for most projects, the benefits of building community and nature connectedness to improve socioemotional health can be realized. We hope our findings can inform the design, and re-design, of CCS projects that may guide project directors in increasing their capacity to respond to and continue through disturbances.
Data Accessibility Statement
Data from the SciStarter database is available via info@scistarter.org, or the authors. Categorized data, R scripts, and interview notes are available via contacting the authors at either sldrill@ucanr.org or cjrosenblatt@ucdavis.edu. Data on COVID-19 restrictions from the following sources: Kaiser Family Foundation (); National Academy for State Health and Policy (), and The Council of State Governments () are publicly available through their websites at the time of this writing.