Investigating Food Type Preferences in Ants: An Experimental Report

This report details an experiment investigating the attractiveness of different food types to ants. The study tested the preferences of ants for honey (sugar), cheese (protein), and cooked rice (carbohydrate) across three different locations: a backyard, outdoors, and indoors. The experiment measured three factors: latency (time to ant appearance), duration (time ants spent at the food), and the number of ants present over an hour. The experiment was replicated over ten days, and a control (water) was included. Statistical analysis using ANOVA was performed to determine significant differences between food types. The results indicated that food type significantly affected the latency of ants, with cheese having the shortest latency. However, there was no significant difference in the number of ants or duration of ants around the different food types. The interaction between the location and food type was significant.

FOOD TYPES ATTRACTIVENESS TO ANTS 1
Food Types Attractiveness to Ants
Name
The Name of the Class (Course)
Professor (Tutor)
The Name of the School (University)
The City and State where it is located
Date

FOOD TYPES ATTRACTIVENESS TO ANTS 2
Abstract
The research was designed to investigate whether different types of food differ in their
attractiveness to ants. The study was guided by the primary question: do different food types
differ in their attractiveness to ants? And this question led to the development of the hypothesis;
different food types have different attractiveness to ants. The analysis of variance (ANOVA)
indicated that honey (sugar), cheese (protein) and cooked rice (carbohydrate) had a statistically
different average latency time (the time it takes ants before they appear around the bait).
However, there was no significant difference in the average duration and number of ants around
the bait between these food types. Further, the interaction between the location and food type
was significant. Thus, it was concluded that food type was associated with how fast the ants
appear around the food. Also, the interaction between location and food type was a significant
factor.
Introduction
This report was designed to test the attractiveness of different food types to ants. In general, the
study was to determine the affinity of ants to different types of food. Three types of food were
put to test on how they impacted the food attractiveness to ants. The three food types used were
honey (sugar), cheese (protein) and cooked rice (carbohydrate). This study will assess three
factors that can be attributed to attractiveness. The first factor is the latency, which measured
how long it took to ants appear in each bait. The second factor was duration, which indicated
how long the ants around the food. The last factor was the number of ants around each bait for a
period of 1 hour.

FOOD TYPES ATTRACTIVENESS TO ANTS 3
In this study, we are determined to answer the question: do different food types differ in their
attractiveness to ants? Thus, the main objective is to determine whether the type of food type is
more attracted to ants. This objective was broken down into the following:
 Does the number of ants differ around the food type in one hour?
 Does the latency of ants differ on the different food type?
 Does the duration of ants around food differ by the food type?
 Is interaction between food type and location of the bait a significant factor on the
number of ants around the bait?
The study objective helped in formulating the research hypotheses which are as follows.
Hypothesis 1:
H0: different food types do not affect the number of ants around the food.
H1: the number of ants is different around different food types differs.
Hypothesis 2:
H0: the food types do not affect the ant latency of ants.
H1: the food types affect the ant latency of ants.
Hypothesis 3:
H0: the average duration of ants around different food type is equal
H1: the average duration of ants around different food type is different.
Hypothesis 4:

FOOD TYPES ATTRACTIVENESS TO ANTS 4
H0: location and type of food and their interaction have no significant effect on the number of
ants around the food.
Ha: location and type of food and their interaction have a significant effect on the number of ants
around the food.
These hypotheses eventually helped in determining whether these three factors measuring
attractiveness were affected by the food type. This hypothesis was based on the fact that there are
different ant species, of which they may be attracted to different food types. For instance, the
sugar ants (Camponotus consobrinus) are mostly attracted to sugary food. Also, we have the
thief ants (Solenopsis molest) which are also referred to as grease ants as they are attracted to
grease. However, this study did not specify the ant species targeted, and hence species of the ant
was treated as covariate factor.
The research predicts that most of the ants will be around the honey, and fewer ants will be
around the cooked rice. Also, an average less time will be observed on the latency around the
honey and cheese. Lastly, the ants are expected to spend more time around the honey food, and
the last time around the cooked rice. More generally, we would expect a significant effect on the
type of food on their attractiveness to ants. The fourth hypothesis was added to determine
whether the interaction between location and food type affect the number of ants around the food
type in one hour’s time. This interaction is expected to be significant.
Materials and Methods
As indicated in the introduction, the main objective was to determine whether the different food
type differs in their attractiveness to ants. Thus, the material and the approach used to attain this
objective is discussed in this section. To test the food type attractiveness to ants, three food types

FOOD TYPES ATTRACTIVENESS TO ANTS 5
were selected; honey (sugar), cheese (protein) and cooked rice (carbohydrate). The baits were
prepared and placed in three locations. One of the bait was placed in the backyard, second was
placed outdoors behind the front door, and third bait inside the living room behind the window.
To ensure randomization of the experiments, five replications were carried out. Each replica was
placed on a three square feet area apart due constraint of factors like house space. The distance
ensured that each replica is independent of each other.
The baits of the three different types of foods were placed approximately ten centimeters apart.
This was a way of making sure that the more the attractive the bait the more the ants will visit
and the more they will hang around the bait. The experiment was repeated for ten days. When
running the experiments, three aspects were recorded; (i) latency, which measured how long it
took for an ant to appear around the bait; (ii) duration, which measured how long the ants were
around the bait and (iii) the number of ants around the bait in one hour. The time, in this case,
both latency, and duration, were measured in minutes and seconds, but later converted into
minutes only. This increased the accuracy of time measurement. Since this was an observational
experiment, there was a need to include a control experiment. In this case, I added another bait
but with water from which rarely ants visited. For that reason, the control was not recorded and
only the three levels of independent variables were considered (honey, cooked rice and honey).
The analysis carried out to answer the research question revolved around testing whether the
average of different groups is equal. Thus, since the independent variable was the type of bait
and had three levels, the most appropriate test for the hypothesis was one-way ANOVA. First,
the assumptions required when running this test were considered a brief summary of the results
will be discussed in the results section. All statistical analyses were performed on R version 3.5.2
(2018-12-20) -- "Eggshell Igloo".

FOOD TYPES ATTRACTIVENESS TO ANTS 6
Results
First, we get the descriptive statistics for the variables. First, the latency average for cheese is
16.225 minutes, which is the least time, whereas that of honey is 22.837 minutes and for cooked
rice is 31.204 minutes. This means that cheese takes less time for the ants to appear on the bait.
On the other hand, the average time the ants take around these baits are 18.48 minutes, 15.687
minutes, and 22.558mintes for the cheese, cooked rice and honey respectively. This indicates
that ants hang around the honey longer than other baits. Lastly, on average, there were 13.70,
12.40, and 16.067 ants around the cheese, cooked rice, and honey respectively. Thus, it was
observed that more ants were around the honey. However, we need to test whether these
averages are statistically different at the level α = .05.
The assumptions required when running the one-way ANOVA include;
 Independence of groups under the independent variable, this was met.
 Independence of observations, which was met since the data events were random.
 Normality; which can be violated and still ANOVA test yield valid results since the test
is robust.
 The dependent variable is on a continuous scale.
 There should be no extreme outliers in the data. We plot the boxplot to determine
whether there was any extreme outlier.

FOOD TYPES ATTRACTIVENESS TO ANTS 7
Figure 1: Combined boxplot for number, latency, and duration
The boxplot indicates that the data did not have very extreme observations. Although the
duration data had some few outliers on the upper side; they cannot be considered as extreme.
The normality test was carried out using the QQ-plot for the three variables. The results are as
follows.
Figure 2: QQ plot for the dependent variables

FOOD TYPES ATTRACTIVENESS TO ANTS 8
The plot for the latency shows a fairly straight line which shows that the data do not deviate
much from the normal distribution. The QQ plot for the duration shows some deviance on both
tails, showing that the data might not be normally distributed, which is the case even for the QQ
plot of the variable number.
Thus, we could run the one-way ANOVA test. We run the first test to test for the first hypothesis
about food type and the number of ants around the bait. The results are as follows.
Table 1: ANOVA summary for number of ants
Df Sum Sq Mean Sq F value Pr(>F)
Food.type 2 207 103.7 1.035 0.36
Residuals 87 8717 100.2
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
The summary shows that there is inadequate evidence against the null hypothesis (F (2, 87) =
1.035, p-value = 0.36). This implies that we should fail to reject the null hypothesis and conclude
that different food types do not affect the number of ants around the food. Thus, there is no need
to perform post – hoc analysis since the averages of the number of ants around different food
type is not significantly different. This test had the power of 0.5941541, meaning that there was a
59.15% chance of detecting an effect if there was one.
We carry out the analysis of the second hypothesis and the results are as summarized below.
Table 2: ANOVA summary for latency of ants
Df Sum Sq Mean Sq F value Pr(>F)
Food.type 2 3381 1690.4 9.764 0.000149 ***
Residuals 87 15062 173.1

FOOD TYPES ATTRACTIVENESS TO ANTS 9
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
The summary shows that there is adequate evidence to reject the null hypothesis (F (2, 87) =
9.764, p-value < .05). Thus, we can infer that the average latency time of ants on different food
type is statistically different. Since there exists a significant difference we carry out post hoc
analysis, and the results are as follows.
Tukey multiple comparisons of means
95% family-wise confidence level
Fit: aov(formula = Latency ~ Food.type, data = food_type)
$Food.type
diff lwr upr p adj
cooked rice-cheese 14.978890 6.878000 23.0797791 0.0000871
Honey-cheese 6.612223 -1.488667 14.7131125 0.1319875
Honey-cooked rice -8.366667 -16.467556 -0.2657771 0.0413166
The summary indicates that the cooked rice and cheese average are statistically different. Also,
the average between the honey and cooked rice is statistically different at the level .05. This test
had the power of 0.9999981, meaning that there was a 99.9981% chance of detecting an effect if
there was one.
Further, we carry out the analysis for the third hypothesis about the food type and the duration.
The results are as follows:
Df Sum Sq Mean Sq F value Pr(>F)
Food.type 2 716 358.2 2.516 0.0867 .
Residuals 87 12387 142.4

FOOD TYPES ATTRACTIVENESS TO ANTS 10
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
In this case, the summary shows that there is insufficient evidence to reject the null hypothesis (F
(2, 87) = 2.516, p-value = .0867). Thus, the duration ants hung around the bait among the three
food types is not statistically different. This test had the power of 0.9354174, meaning that there
was a 93.54174% chance of detecting an effect if there was one.
The last hypothesis was tested and the results are as follows,
Df Sum Sq Mean Sq F value Pr(>F)
Food.type 2 207 104 7.264 0.00125 **
Location 2 7381 3691 258.583 < 2e-16 ***
Food.type:Location 4 180 45 3.150 0.01848 *
Residuals 81 1156 14
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
The summary indicates that there is sufficient evidence that the interaction between location and
food type has a significant impact on the number of ants around the bait (F (4, 81) = 3.15, p-
value = 0.0185). Thus, we can infer that the interaction between location and food type is
important in determining the number of ants around the bait.
Discussion and Conclusion
As expected, in the introduction, the summary statistics indicate that ants spent the longest
duration around honey and the least duration around cooked rice. However, the cheese had the
lowest latency average time, indicating that ants are easily attracted to cheese. On the other hand,
there was no significant difference in the average number of ants around the three types of bait.

FOOD TYPES ATTRACTIVENESS TO ANTS 11
However, when the test of the hypothesis was carried out, the test summary indicated that the
average time the ants take before they appear around the baits is statistically different. The post
hoc assessment deduced that honey and cheese had the least average latency time, which was
statistically different from that of cooked rice. This is to say, the cooked rice had the highest
average latency time, which was statistically different from that of the cheese and honey. Thus,
when storing these types of food, one needs to be more careful, when storing honey or cheese.
That is, one needs to be careful when storing sugary food and protein food.
On the other hand, the time/duration around the different food type was not significantly
different. Which means, on average, the ants spent an equal amount of time on any food type.
Lastly, the average number of ants around different food type in one hour is not significantly
different. The last hypothesis indicated that an interaction between the location and food type
will affect or determine the number of ants around the food. Thus, when storing different food
types considering the storage and the food type is quite important.
For future assessment, one can carry out extensive analysis on the latency, and duration on how
they are affected by different variable’s interaction, such as location, time of the day, and the
species of the ant.