Introduction

The global public health response to the coronavirus pandemic is unprecedented in recent history. Not since the 1918 flu pandemic has a contagious disease prompted worldwide implementation of social distancing policies, including the closing of schools and places of worship, restriction of large gatherings, and limitation of travel. Under the guidance of the World Health Organization and local governments, stay-at-home orders have disrupted daily activity in the majority of the world’s urban centers. During the early months of the pandemic response, 193 countries closed public education institutions (UNESCO 2022). Global movement and tourism decreased, as 91% of the world’s population live in countries that implemented COVID-19 related travel restrictions (Connor 2020). In the United States alone, over 20.5 million people lost their jobs (U.S. Bureau of Labor Statistics 2020). Job loss and transitions to remote work and education drastically reduced local urban traffic, and by the end of April 2020, total vehicle miles traveled had decreased by 60–90%, with greater reductions in densely populated states (Shilling 2020). In the years since the onset of the pandemic, central business districts have not fully recovered in-person workforces or the retail and cultural venues that attract pedestrian traffic (Gibson et al. 2023). Typical temporal patterns in transit and human outdoor presence in cities globally have been altered as a result of the public health response.

A pandemic of this severity has not struck since the advent of megacities in the mid-20th century. In 1900, about 10% of people lived in urban areas, with a total global population of fewer than two billion (Burdett and Sudjic 2007). Today, over 55% of the world’s 7.8 billion people live in cities, and in high-income countries that number exceeds 81% (United Nations 2019). With more people living in high-density urban areas than ever, these environments are distinctly altered by the disruptions caused by social distancing and stay-at-home guidelines. These disruptions are not limited to human society. Human activity is the most influential factor shaping urban wildlife communities (Schlesinger et al. 2008), and changes in human presence may have noticeable effects on the wildlife species that inhabit our cities.

As expanding cities encroach upon natural habitats, the interface between humans and wildlife grows. Urban land cover has increased much faster over the last 40 years compared to historical rates (Angel et al. 2011). Urban land cover is expected to triple by the year 2030 compared to the year 2000 (Seto et al. 2012). Wildlife species face increasing pressure to adapt to life in urban environments. Most species have difficulty adjusting to these novel conditions; thus, urbanization is a major driver of extinction (Pimm and Raven 2000) and a force of biotic homogenization, in which the same commensal organisms now appear in cities globally (McKinney 2006). However, many species have managed to establish urban populations in recent decades, and these populations exhibit distinctly adaptive behaviors relative to their natural counterparts (Minias 2016; Sih et al. 2011; Yeh et al. 2007). Cities around the world are now home to unique animal communities comprising introduced exotic species and urban-adapted native species.

Because humans are the most influential and inextricable part of the urban environment, both exotic and native urban animals have shaped their behaviors around human activity in some form or another. Commensal and urban-adapted species alike may either depend on or take advantage of human-generated trash, restaurant scraps, landscaping, gardens, and feeding stations or other food offerings (Evans et al. 2009; Kumar et al. 2019). These abundant anthropogenic food sources provide the energy to produce enough young to sustain their populations (Chamberlain et al. 2009). Some species use artificial structures for nesting sites, as they resemble the natural structures used in unaltered environments (Wotton et al. 2002). Urban vertebrates are less fearful and experience less stress around humans than their rural or natural counterparts (Abolins-Abols et al. 2016; Blumstein 2006). However, animals in densely populated urban centers may increase their vigilance or flocking behaviors in response to more frequent human disturbance (Valcarcel and Fernández-Juricic 2009). The sudden and widespread reduction of human activity disrupted the survival strategies of some urban animal populations (e.g., Gilby et al. 2021; Soh et al. 2021). Others saw new opportunities arise as disturbances decreased in formerly human-occupied spaces (e.g., Derryberry et al. 2020; Diamant et al. 2023; Walters et al. 2023).

Reports of unusual animal behavior in urban environments circulated on social media platforms since the early weeks of the pandemic response in early 2020. Many of these widely shared stories proved to be false, but not before spreading to millions of viewers. A well-known example is the purported return of dolphins and swans to the canals of Venice, Italy. One such story received engagement with over 1 million social media users before Natasha Daly of National Geographic debunked the reports (Kinefuchi 2020). Social media is a powerful communication tool and may expand the reach of wildlife conservation and education efforts if wielded effectively (Lenda et al. 2020). But the same aspects that allow broad engagement also allow misinformation about wildlife to spread to audiences without expertise to identify it. As climate change threatens cities with more frequent and severe natural disasters (Seneviratne et al. 2012), and zoonotic pandemics similar to the current COVID-19 outbreak are more likely to occur as the urban-wildland interface broadens (Olival et al. 2017), examining and communicating how wildlife responds to disruptions of human activity will help the general public understand humanity’s effects on other species.

Does prior peer-reviewed research exist that can inform our expectations of how urban wildlife reacts to disruptions of the same scope as the pandemic response? We considered catastrophic events dating back to the year 1980, around the time when publications about urban wildlife began to appear in scientific journals consistently (Magle et al. 2012), that might resemble the COVID-19 public health response of widespread disruption of human activity in an urban center. We particularly focused on hurricanes that struck major cities in the United States and two nuclear disasters: the Chernobyl disaster in 1986 and the Fukushima Daiichi disaster in 2011. Additionally, we searched local newspapers for reports of wildlife behavior in response to four recent major hurricanes that struck North American cities: Katrina in 2005, Sandy in 2012, Harvey in 2017, and Maria in 2017. We evaluated if local news sources, more broadly available to the public compared to scientific journals, published stories on wildlife in response to disasters and if newspapers use credible, scientific sources when covering wildlife stories.

Materials and methods

Journal search

To obtain a broad sample of high-impact biology journals, we used Scimago Journal Ranking (https://www.scimagojr.com/journalrank.php) to search the Web of Science 2018 rankings for the highest impact journals in the following categories: (1) agricultural and biological sciences: animal sciences and zoology, (2) agricultural and biological sciences: ecology, evolution, behavior, and systematics, (3) agricultural and biological sciences: miscellaneous, (4) environmental sciences: ecology, (5) environmental sciences: nature and landscape conservation, and (6) environmental sciences: miscellaneous. We combined the top 50 highest impact journals from each category into one list. When fewer than 50 journals were listed for a category, we used all of them. We then removed duplicate entries to create a list of 257 unique journals (Online Resource 1). We sorted these journals by Journal Impact Factor. We searched the top 100 highest impact journals on our list for articles meeting these criteria: (1) the study area must be described as urban; (2) the study must indicate that a sudden reduction of human activity or population has occurred; (3) the study must describe the impact the decrease of human activity had on a wildlife population or community, in terms of either behavior or composition.

We entered each publication title into the Web of Science online search (apps.webofknowledge.com). We set the date range from 1980-01-01 (if the journal existed before 1980) to 2023-06-30. We read all of the article titles listed under each journal title. If an article title indicated that the research may fit our criteria, we read the abstract; if the abstract further supported that the paper may be relevant, we read the full text to confirm if the research met all three criteria.

Targeted search

We determined categories of disruptive historical events that caused reductions of human activity in urban areas and searched for research investigating the impacts of these events on urban wildlife. These event categories were: (1) Natural disasters (e.g., hurricanes), (2) Nuclear disasters (e.g., Chernobyl, Fukushima), and (3) Pandemics (e.g., Spanish flu, COVID-19). We used a targeted keyword search to identify articles meeting our three criteria listed above in all Web of Science journals, regardless of impact factor (Online Resource 2).

Newspaper search

We picked four recent major hurricanes that struck heavily developed urban centers in the United States: Hurricanes Katrina, Sandy, Harvey, and Maria. We chose these events because through our journal and targeted searches, we found that wildlife research was most abundant after hurricane disasters over any other type of catastrophic event. We accessed NewsBank (https://www.newsbank.com) to search local newspapers in the states or territory where each event had the most damaging impact: Louisiana and Mississippi for Katrina, New Jersey and New York for Sandy, Texas for Harvey, and Puerto Rico for Maria. We searched newspapers for reports of wildlife presence or behavior in urban or human-occupied areas following the storm, using the keywords “hurricane” + “[name of storm]” and reading all article titles returned. For each hurricane, we searched newspapers from the date of the storm’s landfall until one year afterward, as we expected most observations to occur in the weeks and months directly after the event. We noted if reports quoted a credible official (state wildlife manager or biologist), professional (private company/nonprofit biologist), academic (university researcher or faculty), or federal (e.g., U.S. Fish and Wildlife Service biologist) source, or if they cited research from a scientific journal.

Results

Journal search

In the 100 highest impact biology journals, we found 13 articles meeting the three criteria outlined above (urban study area, sudden reduced human population or activity level, and relationship to a wildlife population or community). These 13 articles came from eight different journals (Table 1).

Table 1 Articles found in the top 100 highest impact biology journals that meet our three criteria: urban study area, sudden reduced human population or activity level, and relationship to a wildlife population or community. The date range searched was 1980-01-01 to 2023-06-30
Full size table

Targeted search

By using keyword searches, we found 24 articles meeting our criteria (urban study area, sudden reduced human population or activity level, and relationship to a wildlife population or community) from 19 different journals not included in our list of the top 100 highest impact biology journals (Table 2).

Table 2 Articles identified that do not appear in the 100 highest impact biology journals that meet our three criteria: urban study area, sudden reduced human population or activity level, and relationship to a wildlife population or community. The date range searched was 1980-01-01 to 2023-06-30
Full size table

Newspaper search

We found 25 articles in local newspapers describing the behavior or presence of wildlife in urban or human-occupied landscapes in the year following landfall of a catastrophic hurricane (Table 3). The majority of these articles (21 of 25) cited a credible expert source.

Table 3 We searched local newspapers from five U.S. states and one territory affected by catastrophic hurricanes that struck urban centers for reports of wildlife presence or behavior in the wake of the event. We read each article to determine if the author cited an expert source or scientific journal. The majority of articles (84%) cited a credible expert source
Full size table

Discussion

Among the 100 highest impact biology journals we identified, only eight have published articles that meet our criteria: urban study area, reduced human population or activity level, and relationship to a wildlife population or community. Beyond these, our keyword search identified articles from 19 additional journals. Importantly, over two thirds (26 of 37) of these articles are about wildlife during the COVID-19 pandemic, making articles about prior urban catastrophes exceptionally scarce. These numbers are understandably low: large-scale disaster events are rare, arise unpredictably, and accessing affected areas to collect data may be dangerous. Additionally, pre-disaster data may be limited or nonexistent, making comparisons difficult.

Natural disasters

Extreme natural disasters like hurricanes, earthquakes, and tsunamis demand the evacuation of at-risk residents. When natural disasters threaten urban centers, human activity suddenly ceases, similar to the effect of the COVID-19 pandemic shelter-in-place orders. We found that wildlife research during urban evacuations is limited primarily to Hurricane Katrina. The storm occurred in August 2005 and resulted in a one-month evacuation of New Orleans, Louisiana, followed by slow re-urbanization. Because the gulf coast is prized for its many diverse ecosystems, most wildlife studies focused on the impact of the storms on protected natural areas rather than heavily developed areas.

We found nine studies from the aftermath of Hurricane Katrina that considered the effect of human evacuations, among other consequences, on urban wildlife populations. Three belong to a series of publications examining New Orleans’ 10 most common urban bird species (Yaukey 2008, 2011, 2012). These showed that avian diversity plummeted immediately after the storm. The study published in 2008 compared populations before the storm and five to nine weeks after. Yaukey found that the hurricane destroyed natural food sources, such as vegetation and prey species, as well as anthropogenic food sources. Tree squirrels and birds often rely on discarded food and backyard feeders, which were scarce during the first few months after the storm (Yaukey 2008). Comparable effects were produced by the COVID-19 shelter-in-place orders as people refrained from eating at restaurants, diminishing food sources for urban wildlife (Gilby et al. 2021; Soh et al. 2021).

A decade after the storm, New Orleans had recovered only a fraction of its pre-storm population of over 494,000 people; in 2014, that number sat just above 384,000 (United States Census Bureau 2015). From 2015 to 2017, researchers investigated rodent populations of New Orleans in urban areas with different levels of abandonment after the storm (Peterson et al. 2020). They found that areas with higher counter-urbanization (urban areas decreasing in human population and activity) held greater species diversity. These environments had more unmaintained and diverse vegetation, more debris that served as shelter, and less human disturbance. Maximum abundance occurred in moderately developed residential areas, whereas lower abundance occurred in more urbanized and natural areas. This mirrors effects of the pandemic, where non-essential businesses were vacated for months and landscape maintenance may not have been prioritized.

Overall, we found scarce literature on the effect of human absence on urban wildlife populations following a natural disaster. In the case of Hurricane Katrina, few records were kept on wildlife before the storm occurred, so it was difficult to compare pre- and post-storm populations (Gotham and Greenberg 2014). Additionally, it is difficult to differentiate the simultaneous effects of environmental damage and urban evacuation during natural disasters.

Nuclear disasters

We looked specifically for literature regarding wildlife in the evacuation zones of the Chernobyl and Fukushima nuclear disasters. After the Chernobyl nuclear disaster, humans were evacuated from a 4300 km2 area known as the Chernobyl Exclusion Zone (CEZ), located on the border of Belarus and Ukraine. Some articles investigate the impact of the Chernobyl nuclear disaster on wildlife in the CEZ many years later. While most articles we encountered focused on how subsequent radioactive contamination impacted surrounding plant, animal, and human life, a conference abstract by Plenin and Rurv (1995) addressed the response of wildlife to human absence in the CEZ. They discovered that populations of previously hunted birds and mammals increased, while non-game species remained stable. Trees and shrubs grew freely, creating more habitat for wildlife. Consequently, animals that associate heavily with humans disappeared. Dombrovski et al. (2022) examined raptor monitoring data from the CEZ through the lens of rewilding previously human-occupied settlements as a method to restore biodiversity. This article also notes that the vast majority of wildlife research in the CEZ focuses on the effects of radiation exposure; a paucity of research has investigated relief from human disturbance on the population dynamics of the area’s wildlife.

Like Chernobyl, most research on Fukushima’s wildlife investigates the impact of radioactive contamination. Lyons et al. (2020) is the only article we identified that addresses the influence of human evacuations on wildlife after the Fukushima nuclear disaster. The authors observed wildlife occupancy on a gradient of human activity moving away from the center of the disaster, finding that human presence has a greater influence on populations compared to environmental radiation. The authors note, “we know of no Fukushima studies of wildlife that have included humans as a potential confounding variable” (Lyons et al. 2020).

Pandemics

We found no peer-reviewed research examining urban wildlife response during previous pandemics or disease outbreaks. Only during the past half-century have cities begun expanding at such a rapid pace, and it is even more recently that urban ecology has begun garnering widespread attention among researchers (Magle et al. 2012). For example, over 1,000 research articles were published on urban birds in the decade 2006–2015 alone compared to merely 100 articles published before the year 2000 (Marzluff 2017). A pandemic at the scale of COVID-19 has not occurred during this recent era of urbanization and urban wildlife inquiry.

Now a few years out from the onset of the initial public health response to COVID-19, several empirical studies have correlated changes in urban wildlife behavior to reduced human presence. Over two thirds (26 of 37) of publications we identified fall into this category.

Newspaper searches

We expected anecdotal reports of unusual wildlife sightings to appear somewhat frequently in local newspapers following catastrophic hurricanes, similar to social media posts at the beginning of the COVID-19 pandemic. Although only one newspaper article quoted research published in a scientific journal, articles included citations of at least one expert source on wildlife 84% of the time. The most common expert source was a state wildlife official or biologist (48%). Articles were also likely to cite a wildlife professional (28%), followed by a university academic (16%), and finally a federal agency (12%). Many articles (32%) quoted two or more expert sources (Table 3). Online Resource 3 contains a complete list of the newspaper articles analyzed and the type of source cited in each article.

Conclusion

Baseline expectations of urban wildlife behavior in the face of a sudden major disruption to human activity are not currently well-established in published literature. The general public’s interest in wildlife and conservation has grown stronger in recent decades (Unger and Hickman 2020), which may have spurred false and fabricated observations to spread virally on social media. These online platforms prioritize sensational content for distribution over credible or fact-checked content. This trend leads to a misinformed urban populace, and at worst, physical harm to people and wildlife as disruptive urban events become more common. Meeting this desire for urban wildlife behavior-related content is an opportunity for investigators to conduct research in this field, and for journals and newspaper publishers to provide accessible, research-backed data and interpretations.