Learning behavior of the Red Squirrel

Darlenea Copney

Introduction
Mammals are found throughout the Adirondack Park. As all living animals, they behave to survive and have the same basic goal of finding food, water, avoiding predation and reproducing. Behavior plays a key role in how an animal adapts and interacts in its environment. Learning is a behavior that is modified by an experience. All higher vertebrates possess this behavior and it allows them to adapt to the changing environment of the Adirondacks. There are many types of learned behaviors, one of which is classical conditioning. An ongoing study is being conducted to look at the diversity of small mammal communities along a gradient of human impact in the Adirondacks (M J. Glennon, personal communication). To determine the diversity of these small mammals, they were counted by their tracks using an open ended track tube (Drennan et. Al 1998). It is not certain whether one individual small mammal uses every tube in each site or if they are used by multiple individuals.
I predicted that if one individual small mammal can learn to associate food with the track tubes, it could then learn to utilize every tube in a given site. To test this prediction, I studied the learning behavior of the red squirrel (Tamiasciurus hudsonicus) using conditioning techniques at Huntington Wildlife Forest. To achieve this I collected my data using an example from one of the first studies of classical conditioning by the Russian physiologist, Ivan Pavlov (Scott 1958). Classical conditioning is a form of associative learning where an animal recognizes that two or more events are related. However, the animal behaves in response to a substitute stimulus instead of a normal stimulus. The objective of this study was to record and examine learning behavior characteristics with the use of conditioning techniques.

Background
The red squirrel became the source of my study because they were the most actively observed mammals in my study area, which is characterized by a mixed hardwood and conifer forest. Red squirrels feed mostly on pine seeds but their diet also includes, seeds, fruits, nuts, bark, buds, shed antlers, reptiles, insects, tree sap, pine cones, fungi (including mushrooms that are poisonous to humans), eggs, young birds, mice, and young rabbits. They make nests of leaves in a fallen or hallow tree, or a whole in the ground and store many of their seeds and nuts underground, in piles, or in rocks near their home range (usually 1-2.4 hectares). They have a variety of calls to announce home range or presence of an intruder. Vocalizations consist of rattles, screeches, growls, buzzes, and chirps.
The red squirrel can be found in any kind of forest including coniferous, pine, mixed or hardwood forests, and often around buildings. Red squirrels are solitary, diurnal animals that are active throughout the year. Their peak activity times are at dawn and the late afternoon. Red squirrels are very beneficial to the forest ecosystem because of their dispersal of seeds and hypogenous fungi (required by many trees for successful growth). However, they can damage young trees in plantations and crops in storage.

Study Area and Methods
The study was conducted for four weeks in the Adirondack Ecological Centers demonstration area at Huntington Wildlife Forest, located in Newcomb, New York. The dominant conifer species in this area were balsam fir (Abies balsamea), white pine (Pinus strobus), and red spruce (Picea rubens). Important hardwood species were American Beech (Fagus grandifolia). The understory consists predominantly of wood fern and coarse-woody debris. A track tube containing a piece of bait (peanut-butter and oats) was placed three meters from its cache pile and the squirrel was observed nine meters from the track tube. Each observation was made in five-minute intervals for a period of thirty minutes during the morning and afternoon hours. I moved two meters closer to the tube after every use of the conditioning stimulus. A stopwatch was used to obtain the time that the squirrel could be seen (Observational time) throughout the conditioning process. Behavior was recorded in seven different categories: sitting/resting, eating, calling, foraging/gathering, approaching bait, taking bait, and conditioning response.
Two methods of observing conditioning were used. The first method consisted of producing a sound stimulus after a piece of bait was removed from the tube. More bait was quickly replaced after every removal. The process continued up to three times, however after the third removal a sound stimulus was used without replacing bait. The second method was similar to the first with the exception of the sound stimulus after every removed piece of bait. After the third removal a sound stimulus was used to see if the squirrel had learned to associate the stimulus with food. I anticipate that, using the first method, upon taking the bait and hearing the sound stimulus, the squirrel will learn to associate the stimulus with food and produce a conditioned response. When using the second method however, upon taking the bait and not hearing the sound stimulus until the end, the squirrel will not produce a conditioned response. As the number of days of observation increased, I predicted that the squirrel's response time would decrease. Lastly, I predicted that the closer I moved to the tube, the conditioning response of the squirrel would also decrease.

Results
I used the General Linear Model (GLM) procedure for statistical analysis of learning (SAS Institute 1999). I test whether the red squirrel's learning rate would decrease over the number of days it was observed. I then tested to see if my personal distance from the track tube would have an effect on the squirrel's learning rate, as well as the conditioning stimulus. Lastly I tested whether there would be a significant difference in the squirrel's daily activities over the total days it was observed.
There was no change in the squirrel's learning rate over a period of 8 days (P £ 0.221) (Figure 1). My distance from the track tube did not affect the squirrel's learning rate (P £ 0.134) (Figure 2). The use of a conditioning stimulus had no significant effect on the squirrel's learning rate (P £ 0.657) (Figure 3). Lastly, the squirrel's position (sitting/resting, foraging, eating, calling, etc.) over 11 days of observation did not significantly change, as well as the time it spent in those positions (P £ 0.899, P £ 0.388).

Discussion
When observing the red squirrel's learning rate with the use of a conditioned stimulus for 8 days, the red squirrel did not produce a conditioned response. There was no significant difference in whether the conditioned stimulus was used throughout observation or if it was used at the end of observation. The conditioning stimulus was only used 3 times in each trial for the first method; this may not have been enough time for the squirrel to associate the sound of the stimulus with the bait in the track tube. The red squirrels average learning rate in obtaining the bait over the 8-day period had no significant difference. Since I observed the red squirrel at different times during the morning and afternoon hours, those specific times may not have been optimal foraging times, therefore the red squirrel did not visit the track tube as frequently as expected. Results showed that my distance from the track tube did no significantly affect the red squirrel's learning rate. This could have resulted from the red squirrel becoming accustomed by my presence the more I visited the cache pile.
Over the 11 days of observation, the times the squirrel spent resting, calling, foraging, etc. also did not change significantly. Again because I did not observe the squirrel at set times during the day, the squirrel could have spent more or less time doing one thing one day compared to the other. The weather also could have had an effect on the squirrel's activities. A four-week data collection may not be sufficient to draw reliable conclusions about the ability of an individual small mammal to learn to utilize a set of track tubes in a grid. The red squirrel did learn to keep coming back to the track tube to obtain the bait. Therefore the red squirrel did learn to associate the track tube with a food resource.
Consequently there is not enough evidence to determine if individual small mammals learn to go through an entire grid system. Knowing the behavior of small mammals can help understand why there is an abundance of certain species using the track tubes in different sites. This could show that as a whole a specific small mammal species might learn to use the track tubes more efficiently than other species. Learning may be required by different species to solve a given functional problem. Studying the behavior of any animal gives rise to new ideas. Animal behavior can be an effective tool for conducting more effective studies in diversity and management.

Contact Information
If you have any questions, please contact Darlenea Copney at: dscopney@syr.edu


About The Author: Darlenea Copney
Undergraduate student at SUNY-ESF. Research interests include animal behavior and wildlife conservation. After finishing my B.S., I plan to attend graduate school for further studies in animal behavior.




















FIGURE #1                                                                                           BACK






FIGURE #2                                                                                            BACK








FIGURE #3                                                                                         BACK