Responses of urban gray squirrels (Scurius carolinensis) to humans and conspecifics in an area of Boston Common

Emma Rademacher

Read the instructor’s introduction
Read the writer’s comments and bio
Download this essay

ABSTRACT:
Urbanization is rapidly forcing species into different habitats and affecting the way they interact with humans and conspecifics. Studies have explored the large-scale differences between urban and rural species, but have neglected the differences among individuals within populations. This study focuses on the behaviors of a population of Eastern Gray squirrels in the Boston Common Central Burying Ground. I investigated how these urban squirrels responded to humans through habitat selection, boldness, and vigilance, and to conspecifics through social learning and aggression. I hypothesized that the squirrels found in northern areas will be bolder and less vigilant than squirrels found in the southern quadrants. To test my hypothesis, I made qualitative observations regarding the squirrels’ behaviors, then conducted quantitative studies on the squirrels’ boldness and vigilance. To test vigilance, I measured flight initiation distance (FID). To test boldness, I recorded the time it took for each squirrel to obtain an almond from me. My results supported my hypothesis, showing that squirrels were bolder and less vigilant in the northern quadrants than in the southern quadrants. These results can be explained by habituation to humans, density of conspecifics, and canopy cover.

INTRODUCTION

As the world is becoming more urbanized, various species can be found in locations very different from their native rural environments. Prior studies have found that urbanization can lead to heightened neophilia, boldness, and aggression in some species (Barrett et al. 2018). In their 2010 book chapter “Urban Wildlife Behavior,” Amy M. Ryan and Sarah R. Partan explore some of the behaviors of urban animals responding to increased exposure to humans. In many urban species, one of these behaviors can be higher tolerance for humans, as measured by shorter flight initiation distances (FID), or the distance at which an animal changes its behavior as a human approaches it. The Eastern Gray Squirrel is one species whose urban groups have shorter FIDs than their non-urban counterparts (Cooper et al. 2008).

While there are many animal behavior studies like Cooper et al. (2008) which focus on large-scale differences between urban and rural populations, most studies ignore differences in behavior among individuals within a population (Ryan and Partan 2010). In addition, Ryan and Partan (2010) note that the effects of urbanization on social behavior is an area in need of research. This study aims to address these gaps by studying the behaviors of a population of Eastern Gray squirrels inhabiting a section of an urban park in Boston.

Boston Common is a 50-acre urban park in the center of Boston that attracts hundreds of thousands of people every year. Founded in 1634, the Common’s history began as a gathering spot during the Revolution (“Boston Common”). The Central Burying Ground, located in the southern section of the park, is home to the graves of many historical figures as well as a dense population of squirrels that enjoy daily human feeding.

I observed this group of squirrels on three occasions in March 2019 and recorded my observations of their boldness, vigilance, foraging, response to human movement and sound, and social behaviors. In addition, I conducted tests to measure their boldness and vigilance in response to my actions. My study will address how the behavior of urban Eastern Gray squirrels in the Central Burying Ground of Boston Common varies as they respond to human visitors and conspecifics. I predict that most of the squirrels in the Central Burying Ground will be found in the northern quadrants of the cemetery due to habitual human feeding and greater canopy coverage there. The squirrels found in these areas will be bolder and less vigilant, on average, than the squirrels found in the southern quadrants of the cemetery; however, among individual squirrels in the northern quadrants there will be a range of temperaments, with shyer squirrels learning from bolder ones.

METHOD

Study Area
The Central Burying Ground is bordered by Boylston Street to the South, Tremont Street to the East, and Charles Street to the West. This area is home to many Eastern Gray squirrels, as it contains many trees and is protected by a tall fence.

Grassy Triangle
This space is mostly hard dirt closest to the fence, then becomes grassier closer to the walkways. The fence between the Grassy Triangle and the northeast and northwest quadrants is a barrier between the majority of the squirrels and the people, but quite often people feed squirrels there. Many squirrels venture past the fence to be fed by people or to forage on their own.

Northeast Quadrant
The Northeast quadrant is bordered by a walkway on the east side, and a grassy patch on the north side and has a canopy cover of almost 100%. Most squirrels gather here.

Northwest Quadrant
The Northwest quadrant is bordered by a walkway on either side. There is less green space outside the fence surrounding this quadrant as compared to the amount of green space outside the fence surrounding the Northeast quadrant. The Northwest quadrant has a canopy cover of 40%. This area contains the second highest amount of squirrels.

Southeast Quadrant
The Southeast quadrant is bordered by a walkway to the east and Boylston Street to the south. Canopy cover of this quadrant is about 75%. Not many squirrels populate this area. This quadrant has significantly more noise pollution than the Northwest and Northeast quadrants due to its proximity to Boylston Street.

Southwest Quadrant
The Southwest quadrant is bordered by a walkway to the west that contains little human traffic, as well as Boylston Street on its south side. A small gorge separates the majority of the green space from the fence. There is about 25% canopy coverage. There are fewer squirrels.

PROCEDURES

Observations
The first study involved observing squirrels from the Grassy Triangle to the north of the Central Burying Ground. I observed on three different dates in March 2019, one week apart, between the hours of 12:00 p.m. to 3:00 p.m. for periods of 20-60 minutes. Behavior was recorded in a computer roughly every minute and each observation was categorized as Vigilance (V), Boldness (B), Feeding (F) or Other (O).

Vigilance Study
To determine vigilance, I measured flight initiation distance (FID) in a manner consistent with Cooper et al. (2008). I began in the Northeast Quadrant, selected one squirrel at least 10 feet away, walked at a slow, steady pace, with my arms by my side, a blank facial expression, making no sound, with a gaze focused on the squirrel. Once the squirrel turned in the opposite direction and started to run away, I measured the distance between myself and the location at which the squirrel fled, making note of any other behaviors the squirrel exhibited. I repeated this process for another squirrel in the same area, and then for two different squirrels in every other quadrant.

Boldness Study
To measure boldness, I placed myself first in the Grassy Triangle. I sat with almonds in my hand or directly next to me. The goal was to get the squirrel to obtain the nut, whether it were in my hand or beside me, but the way in which the squirrel obtained the nut was taken into consideration when quantifying its boldness. I began a stopwatch when I was able to get a squirrel’s attention. I sat with my legs crossed, waited, and ended the stopwatch once the squirrel obtained the nut. I repeated this process for two squirrels in each quadrant and in the Grassy Triangle. I made note of specific behaviors of each squirrel as it approached me.

RESULTS

QUALITATIVE RESULTS

Qualitative Observations Details Observations were recorded and coded in a spreadsheet which can be found in the file labeled “Squirrel Observation Details.”

Response to Humans

Foraging Locations
Most squirrels present during my observation periods foraged in the northern quadrants of the burial ground and in the Grassy Triangle. About 90% of the total squirrels in the graveyard were in the northern quadrants and 10% in the southern quadrants. When given food from humans, squirrels most often went back through the fence to eat or bury the nut inside the graveyard. When foraging on their own, most squirrels stayed within the northern sections of the graveyard; only a few squirrels would occasionally venture outside the graveyard to forage in the Grassy Triangle. When a person left a pile of nuts on the ledge surrounding the burial ground, usually no more than three squirrels at a time gathered around the pile.

Vigilance
Squirrels responded with greater vigilance to humans that were noisier and made quick, large movements. A man trying to feed a squirrel in the Grassy Triangle right outside the Northeast Quadrant was talking loudly and continuously. The squirrel he was trying to feed was cautious in approaching him—it would approach a few feet, then freeze on all fours when the man tried to move closer to it. After three minutes, the squirrel took the nut from the man’s hand.

When I sat quietly on the Grassy Triangle between the Northeast and Northwest Quadrants, some squirrels would pause on all fours to stare at me from about 15 feet away on the ledge surrounding the burial ground. When I shook the bag of almonds to get a squirrel’s attention, a few squirrels would either freeze or flee. A few of the squirrels flicked their tails and stared at me if I got too close, made too loud a noise, or made too quick a movement.

Boldness
Overall, squirrels responded with greater boldness when I sat still and made quiet but repetitive kissing or clucking sounds. I observed that most squirrels also responded positively to the shaking of a bag of almonds, followed by an outstretched hand. If I sat still outside the graveyard in the Grassy Triangle with almonds in my hands about 95% of squirrels would approach me and take a nut from my hand. Squirrels in the Grassy Triangle and the northern quadrants were quicker to take nuts from my hands than squirrels in the southern quadrants.

Squirrels were unafraid of human possessions, such as my backpack and laptop. As I was typing notes, about one to two squirrels every two to three minutes would approach my laptop. While I was paying no mind to the squirrels and typing up observation notes, a few squirrels crawled under my crisscrossed legs to stick their heads into the bag of almonds.

Response to Conspecifics: Social Learning and Aggression
On several occasions, onlooking squirrels appeared to copy the behavior of bolder conspecifics. When I sat still, a squirrel would cautiously approach my hand, but eventually obtain the nut. Another squirrel nearby observed this behavior and approached me, looking for another nut, only a few seconds afterwards. While sitting in the Grassy Triangle, I observed similar behavior, even without nuts.

A sudden abundance of food in any particular area created aggression amongst the squirrels. Within the northern quadrants, squirrels would often chase each other away while foraging. Similarly, if a human placed a pile of nuts on the ledge surrounding the burial ground, about three squirrels would approach the pile, and they would each try to chase the other away so that they could return to the pile and eat by themselves.

QUANTITATIVE RESULTS

Vigilance Study Results

My results indicate a significant difference in FID between squirrels in the northern and southern quadrants.

Figure 1: Vigilance Study Results from 3/31/19

Boldness Study Results

My results indicate a significant difference in average time for squirrels to obtain a nut between the northern and southern quadrants. In the northern quadrants, squirrels took, on average, 36.25 seconds to obtain a nut. In the southern quadrants, squirrels took 57.5 seconds to obtain a nut, on average.

Figure 2: Boldness Study Results from 3/31/19

DISCUSSION

Habitat/Foraging Location
My study contributes to the existing literature of how habitat suitability and human habituation impact the foraging and habitat selection of urban Eastern Gray squirrels. My results indicate that, despite their proximity to heavy human foot traffic, the Northern quadrants of the Central Burying Ground are a preferred foraging location for squirrels in Boston Common.

Several prior studies provide possible explanations for this preference. A 2008 study found that canopy cover is a crucial determining factor for where squirrels choose to reside (Parker and Nilon 2008). Another study suggests that squirrels can habituate to humans and associate them with food. The risk-allocation hypothesis (Cooper et al. 2008) predicts a decrease in anti-predator behaviors in areas with high levels of human activity (discussed below), and an increase in habituation towards human feeding. Furthermore, while repeated exposure to the sights and smells of humans can lead to either sensitivity or habituation towards humans, species living in urban environments are more likely to become habituated to humans rather than avoidant of them (Barrett et al. 2018).

These studies support the claim that the Central Burying Ground squirrels have chosen the northern foraging location for its canopy coverage and due to their habituation toward human feeding. However, since the current study area has several features which seem to make it desirable for squirrels, it is not possible to isolate the impact of each feature. A future area of research could be how human structures and barriers such as the iron fencing around the graveyard impact gray squirrel foraging location selection.

Response to Humans: Vigilance and Boldness
While most studies of urban gray squirrels compare their vigilance and boldness behaviors to those of their rural counterparts (Barrett et al. 2018; Nowak et al. 2018), the current study takes a closer look at the vigilance and boldness behavior within a single group of urban squirrels. My results show that within the Central Burying Ground group, vigilance behavior varied by location in the graveyard with squirrels in the north exhibiting greater boldness and less vigilance than squirrels in the south. Previous studies on human habituation, conspecific density, and canopy coverage offer several possible explanations for this variance. Future studies would be needed to isolate each of these variables to determine which has the greatest impact on boldness and vigilance.

Human Habituation
One possible explanation for the variance in behavior could be that the squirrels in the Central Burying Ground are behaviorally flexible in response to varying conditions in each part of the graveyard. This may be because they have learned through experience that humans in the north are generally not threatening and are sources of food, whereas they have less experience with human behavior in the southern part of the graveyard. A study by Nowak et. al. (2018) found that urban squirrels have lower giving-up densities ( GUDs) than their rural counterparts, suggesting that the more exposure squirrels have to humans, the less wary they become in their presence. In addition, a study of rainbow trout sought to determine whether their behavior was consistently bold or shy (i.e., “domain-general,” as in humans), or whether their behavior was context-specific. The study found that shyness and boldness in rainbow trout depended on context (Wilson and Stevens 2005). Perhaps the squirrels in the current study have learned to associate comfort around humans with the specific location in which it occurs, and do not generalize about all humans in all locations. However, further study of the squirrels would be needed to determine whether the same squirrel behaves differently in different parts of the graveyard, or if it is uniformly bold or shy throughout the cemetery.

Density of Conspecifics
Another explanation could be that the squirrels are bolder and less vigilant when there are higher densities of squirrels, such as in the northern quadrants. A previous study suggests that when squirrel density increases, wariness decreases and intraspecific boldness and aggression increases (Parker and Nilon 2008), supporting the idea that the higher density of squirrels in the northern quadrants would be cause for higher levels of aggression there.

Canopy Coverage
The squirrels’ variance in boldness and vigilance behavior can also be explained by the amount of canopy coverage in each section. Since the canopy coverage is much less in the southern quadrants, the squirrels may have left more distance between humans and themselves because of greater distance from a tree, perceived by squirrels as a guaranteed escape route. Previous studies have supported this idea, stating that squirrels have positive associations with high levels of canopy coverage, as it provides shelter and safe places to forage (Shuttleworth et. al. 2016). Parker and Nilon also found that canopy coverage is a major predictor of squirrel wariness as well as squirrel population density, thus also explaining why squirrels congregate more in the northern quadrants. They claim that more canopy coverage allows squirrels to be less wary, as it provides protection against predators (Parker and Nilon 2008). The results of these studies and the current study ultimately suggest that greater canopy coverage allows for squirrels to be bolder.

Response to Human Movement and Sound
The current study contributes to the existing literature that shows varied animal responses to human movement and sounds.

Movement and Gaze
Overall, squirrels were bolder when I sat still and did not engage in direct eye contact with them. Previous studies offer possible explanations for the squirrels’ inquisitive behavior. Barrett et al. (2018) claim that bolder individuals are more likely to become habituated to the presence of humans, also making them more likely to engage in riskier behaviors, such as stealing anthropogenic food. It is possible that the squirrels have learned that certain body movements or direct eye contact indicates a threat. Similarly, Marzluff (2010) found that crows rely on cues from conspecifics and heterospecifics to gather information about threatening humans (Barrett et al. 2018). In addition, Zou et. al. (2014) found that monkeys associate direct eye contact with an unfamiliar human as threatening.

Sound
My results also showed that the squirrels in the study area were attracted to some human-made sounds and repulsed by others. Human kissing or clucking sounds attracted bolder squirrels, and even encouraged squirrels that were exhibiting the highest degrees of vigilance. This could be because the squirrels associate humans that make these noises with sources of food. Each time I made a kissing sound, I followed through with an outstretched hand and a nut.

However, other sounds, such as loud talking or laughing, deterred the squirrels. One such instance includes the squirrels’ interactions with the noisy man standing on the Grassy Triangle outside the northeast quadrant. According to prior studies done by Levey et. al. (2009) and Vincze et al. (2015), discrimination learning allows species to avoid particular humans with whom they have had unpleasant experiences in the past (Barrett et al. 2018). Perhaps the noisy man had been to the cemetery before, causing a commotion, and leading the squirrels to believe he was a potential threat. Belguermi (2011) and Stephan (2013) also identified several studies in which some species can identify and remember particular humans by their facial features or by their particular behaviors (Barrett et al. 2018).

Other human-made sounds could be affecting the squirrels in the cemetery. Lower squirrel density in the southern quadrants could be attributed to higher levels of noise pollution, since the southern quadrants are bordered by Boylston Street. Similarly, Duarte et. al. (2012) found that urban marmosets avoid areas with heavy sound pollution (Barrett et al. 2018).

Response to Conspecifics: Social Learning and Aggression
This study also contributes to prior research on the social behavior of urban animals foraging in groups.

Social Learning
My results showed that shyer on-looking squirrels seemed to learn from bolder squirrels how to safely obtain food. These results add to previous studies which show social learning among urban squirrels and other species.  One study found that red squirrels can learn new feeding techniques from a more experienced squirrel and that this new knowledge persisted even after the model squirrel was no longer present (Weigl and Hanson 1980). However, another study found that squirrels learn even more effectively when observing a conspecific fail, rather than succeed, at a task (Hopewell et al. 2009). Hopkins (2013), Mazur and Seher (2008), and Breck et al. (2008), have done studies of other species including the black bear and jackdaws, which have found that these animals also learn from each other how to forage on anthropogenic food (Barrett et al. 2018). Future studies could investigate whether bolder squirrels, like the ones in the current study doing the “teaching,” are the parents of the shyer individuals, or whether shyer adult squirrels typically learn from bolder adult conspecifics.

Aggression
My results showed that intraspecific aggression increased in an area of high squirrel density when anthropogenic food was introduced. This finding is consistent with a study by Parker and Nilon (2008) which found an increased squirrel density contributes to decreased wariness and a more competitive drive for survival, thus creating more opportunity for aggression amongst squirrels. However, another study found that an increase in squirrel density leads to a decrease in squirrel aggression (Haigh et al. 2017). Since my study added the variable of anthropogenic food, this could explain the differing results. Future studies could examine the impact of density and human feeding on squirrel aggressiveness with a larger sample size of squirrels. In addition, future work could expand on this research by exploring the levels of aggression in squirrels when the density of humans or heterospecifics increases or decreases.

CONCLUSION

Limitations
There were various aspects of my study that may have been cause for error. I was inconsistent in how I executed the boldness study. For some squirrels, I held the nut in my hand, while for others, I placed the nut on the ground beside me. Furthermore, I made repetitive kissing and clucking sounds for some squirrels, but not for others. Another point of error in my study could have resulted from too few trials. My results would need a larger sample size to be statistically significant. Additionally, since I returned to the cemetery three times and stayed for several hours each time, the squirrels may have habituated to me, and grown bolder in the process. Finally, when I quantified the number of squirrels that appeared bold or vigilant, I may have recounted squirrels.

Areas for Future Research
Future studies could address the limitations above through larger sample sizes, more trials and more standardized procedures for interacting with the squirrels. Additionally, squirrels could be tagged to determine whether they are consistently bold or shy, or whether their behavior is context-specific. Future projects could also attempt to isolate human habituation and canopy coverage to determine which has the greatest impact on boldness and vigilance. And finally, future studies could investigate whether bolder squirrels, like the ones in the current study doing the “teaching,” are the parents of the shyer individuals, or whether shyer adult squirrels typically learn from bolder adult conspecifics.

The current study shows a range of behaviors within a small group of urban squirrels and suggests that these differences may be due to fairly subtle changes in the landscape and in human-animal interactions. Studies on individual urban populations are important because they are a starting point to help us enhance conditions for both humans and wildlife in specific urban locations.

References

This article is a review of the special cognitive abilities of species which typically come into conflict with humans and are therefore called a “nuisance.” These abilities include boldness, categorization, innovation, memory, learning, social learning and behavioral flexibility. This article will give me good definitions of behaviors and help me locate other studies which address vigilance and boldness. The sections on boldness, learning and social learning will be helpful to understanding the possible relationships between habitual human feeding and squirrel cognition, temperament and behavior.

This site provides general information as well as a brief history on Boston Common. I used this source for background on Boston Common in my introduction.

In this experiment, a mechanical robot squirrel was programmed to mimic squirrel tail flags and alarm barks so as to test squirrel responses in both urban and rural environments. The study found that the squirrels responded most when both tail flags and alarm barks were combined rather than isolated, and that urban squirrels were more responsive to the tail flags, suggesting that the noise of an urban environment might be causing a shift to multimodal or visual cues rather than solely acoustic ones. This study is useful background to aid in the understanding of the effect of urbanization on squirrels.

In this chapter of the book Urban Wildlife Conservation: Theory and Practice, authors Ryan and Partan outline many of the ways in which animal behavior changes due to urbanization. These areas include animal movement and home ranges, use of human structures, foraging behavior, anti-predator behavior and response to humans, social behavior, and animal communication. The authors conclude with a look at behavioral flexibility and the evolutionary implications of the effects of urbanization. This article is useful in my project because the sections on anti-predator behavior and response to humans, as well as on behavioral flexibility, reference several studies which can help me determine my specific focus and provide background information.

Habitat Selection

In this book chapter, the author looks at the ecological characteristics of urban and rural squirrel populations, ecosystem functions and services of urban gray squirrels, the ecological impacts of urban gray squirrels on native fauna, and the management and control of urban gray populations. The authors consider it important to study urban gray squirrels to allow predictions of their behavior in new urban environments and to allow development of effective wildlife management strategies. This chapter is relevant to my study of Boston Common squirrels because it examines the impact of different habitat features, such as canopy coverage, on population density.

This paper explores the relationship between between gray squirrel density, intraspecific aggression, and boldness and habituation to human presence in various parks. This article measured these variables in Lafayette Park in Washington D.C., as well as in six other urban parks in Baltimore, MD. Ultimately, the study found that there is a positive relationship between squirrel density and intraspecific aggression, and a negative relationship between density and wariness. This research is helpful for my current study because it will provide support for my hypothesis.

Response to Humans

In this experiment, squirrels were given trays of sunflower seeds at seventy-eight sites in Virginia. Giving-Up Density, or GUD was measured by counting the number of seeds left after a period of time. Researchers found that the lowest GUDs occurred in urban areas near human structures, suggesting that either squirrels have reduced predatory fear of humans or that they have limited access to food in urban areas and are more hungry and willing to expose themselves. They also had lower GUDs when there were higher squirrel densities and greater canopy cover, suggesting that the squirrels fear predators less when there are more of them and they have an easy escape route up a tree. This study is relevant to my study of vigilance and boldness of squirrels.

This article written by Cooper et. al explores the responses of squirrels when approached by a human only and a human with a dog in areas with high and low levels of human activity. The authors conclude that alert distance in squirrels overall does not vary between human alone and human with dog approaches, but anti-predator behavior is influenced by the presence of humans in the area. This article will be helpful as I am also studying vigilance behavior in the squirrels on Boston Common.

This article examines the levels of shyness and boldness in squirrels at various locations, including urban, forest, and rural environments by measuring Giving-Up Density (GUD). The research found that squirrels in open environments were more likely to experience higher GUDs, and that urban squirrels experienced the lowest GUDs across distances of 5, 10, and 15 feet. The report attributes the cause of this to the high level of human activity in any environment would make squirrels more habituated to the presence of humans, or even make them see humans as a source of food. This research is relevant to my paper because I am also examining the level of boldness and vigilance in squirrels in an urban environment. This article will help me support my hypothesis.

This study asks whether rainbow trout are consistently bold or shy (domain-general) or whether their behavior is context-specific. The authors found that bold trout were consistently bold when the context was foraging, but that when exploring a swim flume, their behavior varied. The authors concluded that rainbow trout shyness and boldness depends on the context they are in. This research is relevant to my study of squirrel boldness and could also provide ideas for future methods of squirrel boldness/shyness research.

In this experiment, monkey behavior was analyzed when a human observer gazed at them and looked away. Various response behaviors were recorded such as opening/closing mouth, agitation, staring back and approaching the observer. The study found that the monkeys exhibited heightened awareness while being gazed at by the humans. They had longer durations of the measured behaviors when being gazed at. Similarly, during my study, the squirrels at the Central Burying Ground reacted with vigilance or fear when I fixed my gaze upon them.

Response to Conspecifics

This study used radio tracking of squirrels and examined the effects of squirrel density on activity and aggression, survival probability, breeding and body condition. The researchers found that in areas of high squirrel density, the squirrels were less aggressive than their counterparts in areas of low density. There was also a significant correlation between the activity level of the squirrels and their aggressiveness. This study is relevant for my project because I too am looking at the relationship between squirrel density and aggressiveness.

In this study, researchers found that eastern gray squirrels who watched a conspecific fail at obtaining a food reward learned more readily than those who watched a conspecific succeed. My observations indicate that shyer squirrels become more bold as they watch bolder conspecifics obtain food from humans. Since these results seem to contradict each other, I will make note of this in my discussion. This study can also provide ideas for future experiments on social learning among Boston Common squirrels.

In this study, researchers studied the amount of time it took red squirrels to learn how to eat hickory nuts for the first time. One set of squirrels used trial-and-error, while another set was able to observe an experienced conspecific feeding on the nuts. This study found that the squirrels who learned from the model squirrel were more efficient than those who learned via trial-and-error, and that this learning persisted even when the model squirrel was removed. This study supports my claim that social learning occurs among squirrels in the Central Burying Ground.