Ethogram of Pogonomyrmex occidentalis

Introduction
For this ethogram I observed the species Pogonomyrmex occidentalis, more commonly
known as the western harvester ant. These ants live in the dry grasslands and deserts of
the American West. Indiana is not a desert nor grassland, so I was unable to watch them
in the wild. I purchased an ant farm that is based off a farm NASA scientists use to test
ant behavior in space. It is filled with a thick, blue gel that contains all the necessary
nutrients and water an ant may need to survive. Before observing the behavior of the
ants I allowed them 3 days to acclimate to the environment of the farm. I would have
liked to observe an ant species native to Indiana however it is difficult to watch their
behavior when much of it is done underground or in their mounds. The western
harvester ants were a good choice for me because they have larger mandibles than the
typical ant which allowed them to tunnel through the gel more easily.

Once the ants became familiar with the environment and started creating their tunnels
in the gel, I began observations for the behavior cataloging and time budgets. Given
that the ants were in such a confined space I knew that diversity of behavior would be
limited. To name the behaviors I simply watched them and marked anything I noticed
them doing. That included simple individual behaviors or interactive behaviors. All the
behaviors I noticed through observation are listed below. Through observation I saw
that some ants were much more active than others and I wanted to account for this in
the time budget. I utilized a scan sampling method for this and focused on two ant
types, a very active ant, and an inactive ant. Knowing that behaviors are very
individualized I wanted to look at the difference between the two extremes, showing
lots of behavior and showing little to no behavior. I observed these ants with my naked
eye, three days in a row between the times of 3 and 4pm.

Ethogram

Inactivity

Standing or crouching
in one area with
Stationary: antennae flinching
periodically.
 Activity

Mandible is used to
chew and break down
Eating: gel to absorb its
nutrients.

Walking up the farm

walls slowly with all
Climbing: legs spread out widely.

Using mandibles to rip

gel pieces away from
Digging: the farm creating
tunnels.

Walking forwards
slowly with antennae
tapping the gel in front
Exploring: of them repeatedly and
in all directions.
 Two ants facing each
Communicating: other tapping their
antennae together.

Moving pieces of gel

Organizing Waste: waste from digging
tunnels into a pile.

Dragging antennae
Antennal underneath the
Grooming: frontmost legs
repeatedly.

A combination of
carrying and dragging
the carcass of another
Funeral: ant into an area with
waste.
Time Budget
To conduct my time budget I utilized a scan sampling method. I set a recording device
on the ants for 15 minutes, three days in a row for a total of 45 minutes of recorded
behavior. I then split those 15 minutes up into 20-second intervals for a total of 252 data
entries, 126 for each ant type. At each 20 second interval I would make a mark of which
behavior the ant was entertaining. The table below shows how many time periods each
ant type spent performing one of the behaviors of interest. Just from looking at this
table we can see a few things. Not a single ant participated in the funeral behavior, there
was also very little interest in organizing waste. The two most common behaviors I
observed during sampling were stationary and exploring. The digging behavior was
also of interest to me because every active ant participated, and no inactive ants did.
Given that the overall purpose of these observations was to describe behaviors elicited
by P. occidentalis I will not be going into depth about the behaviors displayed by the
different ant types.

Looking at the figures, Table 1 allows us to visualize the data collected. Each behavior is
listed on the top and the numbers represent how many total time periods were spent
doing that behavior in a sampling period. Just from this table we can see that the most
common behavior amongst the ants was exploring, with a total of 71/252 observations.
We can also see that the stationary behavior was much more common with the inactive
ants than the active ants. As well as digging being more common in active ants than
inactive. Overall, this table shows a good distribution of each behavior taking place
during the sampling periods aside from the funeral behavior.

Figure 1 is showing a stacked bar graph displaying at which proportion each ant type
spends eliciting the named behaviors. Again we can see a much more diverse behavior
pattern in the active ants than the inactive. If we look at the most common behaviors
elicited when combining the data for both ant types, we can see that they spend 28.2%
of their time exploring, 33.3% of their time stationary, and 13.5% of their time digging.
The remaining 25% of their time was spent participating in the other behaviors such as
eating, climbing, communicating, organizing, and grooming. Although the funeral
behavior was seen during ethogram observations it was not seen during the time
budget sampling accounting for 0%.

Table 1. A visualization of all data collected showing both ant types and each behavior
they participated in.
 Percent of Time Spent Eliciting Behavior
100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

0%
Active Inactive

Funeral Grooming Organizing

Communicating Exploring Digging
Climbing Eating Stationary

Figure 4. Stacked bar chart displaying what percentage of collected data points ants
spent engaging in each named behavior.

Challenges and Limitations

Regarding any limitations within this study I believe that I gave myself an upper hand
by housing the species I was observing. I was able to observe the exact same species, in
the exact same environment, and at the exact same time for every sampling period. I
was also able to utilize a recording device to ensure that the inactive ants and active
ants were being sampled at the same time. A challenge that was present was that I was
unable to observe the same ants each day I sampled data. I had about 20 ants in the
farm and was unable to tell the difference between them. To try and make up for this
anytime I was collecting data for the time budget portion I would look for the extremes
of each ant type and observed those individuals.

I also experienced some limitations when deciphering between the digging, eating, and
organizing behaviors. I observed the behaviors of these ants with the naked eye and all
though they were larger ants they were still small. At some points it was difficult to
determine was exactly they were doing with their mandibles. Whether they were head
down into the gel for nutrition or digging purposes and whether they were holding gel
pieces in their mandibles for nutrition or organization. To try and prevent some of this I
did not begin time budget sampling until a tunnel was built large enough for the ants to
crawl in and out of. At this point they had also created hills of gel waste which allowed
me to make more educated assumptions about which behavior was being elicited.
Hypothesis
The ant farm that I set up for my observations consisted only of worker ants. Queen
ants are mostly responsible for populating the colony however I would like to know if
she has any effect of the activity of ants. I was very surprised at how many ants in my
farm were inactive and stationary. My hypothesis is as follows; Ant colonies benefit
from having a queen ant that provides structure and the promise of repopulation
therefore increasing their worker activity and behavior. My prediction for this
hypothesis is if an ant colony has a queen all the ants will display a more diverse
behavior profile than they would in a colony without a queen. To do this would be
quite simple and it would not deviate from my previous methods at all. I would
purchase a new ant farm and a new colony, but this time introduce a queen. I would
allow them to acclimate and observe their behavior and sample them for the time
budget in the same way. Aside from just testing that hypothesis it would also be cool to
specifically do an ethogram on the queen and observe her behavior as she is at the top
of the hierarchy and most definitely displays different types of behavior from the
typical worker.