Analysis of Squirrel Activity Patterns and Fur Color Relations in Central Park


Tammy Zhang tz332, Zhuoya Li zl928, Florence Li hl2472, Joey Kassenoff jdk333


The 2018 Central Park Squirrel Census dataset offers a unique glimpse into the lives of eastern gray squirrels (Sciurus carolinensis) in New York City’s Central Park. Compiled by The Squirrel Census through the efforts of hundreds of volunteers, this comprehensive dataset encapsulates a wide array of observations from 3023 squirrel sightings. Each row in the dataset represents a distinct sighting, capturing details such as the time and location of the observation, the age and fur colors of the squirrel, along with various behaviors exhibited during the sighting. The dataset is rich with 31 different columns, providing a multifaceted view of these urban squirrels’ daily lives, interactions, and activities within their environment.

Question 1 : How does the time of day (shift) influence squirrel activities and interactions?


This analysis seeks to explore how different times of the day affect the behaviors and human interactions of eastern gray squirrels in Central Park. The question is of particular interest as it may reveal patterns in how these urban wildlife creatures adapt to their environment and the daily rhythm of the park’s human visitors. The primary focus will be on the “Shift” variable (AM/PM) and observed behaviors such as Running, Chasing, Climbing, Eating, Foraging, Approaches, Indifferent, and Runs from. By examining these aspects, we aim to uncover any significant differences in squirrel activities and interactions with humans between the morning and afternoon shifts. Understanding these patterns can provide insights into the behavioral ecology of squirrels in an urban setting and inform park management and wildlife observation strategies.


To investigate how the time of day affects squirrel activities and interactions in Central Park, we will create two types of visualizations using ggplot2: a bar plot and a histogram. The bar plot compares the proportion of observations including various activities (Running, Chasing, Climbing, Eating, Foraging) between morning (AM) and afternoon (PM) shifts. This visualization is ideal for categorical comparisons since encoding values through length is most intuitive for perceiving differences, allowing us to directly contrast activity frequencies across times of the day. By using color mapping to differentiate between shifts, the plot clearly illustrates which activities are more prevalent in the AM versus the PM, offering insights into temporal behavior patterns of the squirrels.

The histogram, on the other hand, focuses on the distribution of a “Total Activity Index,” a composite measure that integrates various activities and interactions with humans. Except for the 5 activities included in the first plot, we will also take quaas, kuks, and moans into account. This plot type is adept at showing the spread and central tendencies of quantitative data, making it suitable for analyzing the overall activity levels of squirrels across shifts in a day. Through color differentiation and bin width adjustments, the histogram reveals variations in activity intensity during different shifts, providing a comprehensive view of how squirrel behavior changes from morning to afternoon. Together, these plots enable a nuanced understanding of squirrel activities in an urban environment, highlighting the influence of time on wildlife behavior in a manner that is both accessible and visually engaging.


`summarise()` has grouped output by 'Shift'. You can override using the
`.groups` argument.


The first bar graph illustrates the proportions of observed squirrel activities in Central Park, differentiated by time of day with AM in red and PM in blue. The activities are categorized as Chasing, Climbing, Eating, Foraging, and Running.

From the graph we can note that foraging and eating in the PM both show a significant increase compared to the AM, possibly due to squirrels taking advantage of quieter park periods later in the day or increased food availability from park visitors. Climbing is observed more in the AM, which might be due to the squirrels’ morning routines. Running and chasing were markedly similar between the AM and PM, suggesting that these behaviors are observed at even rates throughout the day. The differences between AM and PM activities may be influenced by various factors, including the squirrels’ biological rhythms, human park traffic, and the availability of resources throughout the day.

The second histogram displays the “Distribution of Total Squirrel Activity Index” categorized by time of day with morning (AM) in red and afternoon (PM) in blue. The total activity index, compared with the activity categories above, is a more comprehensive measure that take more activities and human interaction into consideration. The PM (afternoon) activities, represented in blue, have a consistently higher count across most of the activity index categories compared to the AM (morning) activities, depicted in red. This trend indicates that squirrels in Central Park are more active in the afternoon across a range of behaviors that contribute to the activity index. Possible explanations for this could include environmental factors such as temperature, which may be more conducive to squirrel activity in the afternoon, or park dynamics, such as increased availability of food due to higher human traffic later in the day. We can see that most observed squirrels have a total activity index between 0.5 and 1.5. This peak at a lower activity index suggests that there is a common level of observable activity among the squirrels, with a drop-off as the activities become more vigorous or less common. The data implies that while squirrels are active throughout the day, their most detectable behaviors tend to occur during the PM shift.

Question 2 : Is there a relationship between squirrel fur color and their location or activities?


The relationship between squirrel’s physical characteristics, such as fur color, and its behavior or habitat selection can offer insights into how they adapt and survive in the wild. This question aims to explore such a relationship in squirrels by analyzing two aspects of their ecology: location preference and activity patterns in relation to their fur color. The dataset required to answer this question must include variables for fur color, location, and observed activities of squirrels. Investigating this relationship is intriguing as it may reveal whether certain fur colors are more adaptive in specific locations or if they influence the type of activities squirrels are more likely to engage in, perhaps because of how easily they can be seen by predators.


The bar chart is used to compare the average frequency of different activities (Climbing, Eating, Foraging) among squirrels with different fur colors in the second plot. Each fur color has a set of bars representing the activity frequencies, with different colors distinguishing between the activities. This plot is ideal for summarizing categorical data and allows for immediate visual comparisons of magnitude encoded through length. The choice of a bar chart is suitable for showing average values of a numeric variable across different categories and can indicate if certain fur colors are associated with higher or lower frequencies of specific behaviors.

Two tree maps are used for the second graph to represent the proportion of squirrels with different fur colors within two distinct locations: Above Ground and Ground Plane. Tree maps are beneficial when we want to show the composition of a dataset as parts of a whole, which makes it straightforward to see the proportion of each fur color within a specific location.



The first plot shows the average frequency of three different activities (Climbing, Eating, Foraging) for each primary fur color. There seems to be a trend where the cinnamon fur color category has the highest frequency for eating and foraging activities, while black has a noticeably lower frequency for these activities but a higher frequency for climbing. The higher eating and foraging frequencies in cinnamon squirrels might be indicative of their habitat preferences or physical abilities.

The second plot suggests that the consistent majority of gray fur in both locations could indicate that gray is the most common fur color among squirrels in the observed area, possibly due to genetic factors or a selective advantage that makes gray squirrels more likely to thrive. The smaller proportions of black and cinnamon squirrels suggest that these fur colors are less common in the population, which could be due to a variety of ecological or evolutionary reasons. It also appears plausible that black squirrels are observed more frequently above ground than on the ground plane, which is in line with the other plot, which expressed black squirrels are observed climbing more relatively frequently.


Our presentation can be found here.


The Squirrel Census. (2018). 2018 Central Park Squirrel Census - Squirrel Data, United States, NYC OpenData.


Our dataset source: