Tiring Termites Give Up Rather Than Slow Down: Insights to Metabolism Through Rate of Travel

Tiring Termites Give Up Rather Than Slow Down: Insights to Metabolism Through Rate of Travel

Paul Fischer


Ravi Nagori, John Mitchell


Tiring Termites Give Up Rather Than Slow Down: Insights to Metabolism Through Rate of Travel


Discussion has been given to the nature with which termites travel using bridges and pheromones as factors in rate and distance of travel. Limited laboratory equipment and resources as well as an exhaustive number of experiments will require significant effort to eliminate redundancy or inaccuracy in the trial which will be provided. After preliminary observations which identified some key characteristics in the nature and travel of the termites, a novel approach to determining existential factors in the metabolism of the insects has been concluded.

Some mention of metabolism is provided in the research from Loreto et al. (2013) but statistical analysis was not able to ensure a consistent nature to the basic rate of travel of termites, and in one trial results were quite unexpected. This offers some basis for return to laboratory experiments and investigation of possible sources of the inconsistent results. The experiment will evaluate the rate of travel at two points in a track by insects after eliminating outlying results, which will hypothetically establish the presence of a “tiring” mechanism, or a metabolism factor in the termites. This will be shown with a significant deviation (p<.05) between mid-trial termite times and terminal termite trial times. Rejection of the hypothesis will indicate that termites travel at a consistent pace throughout the course of the trial.


Observations were completed using termites in wet, infested logs. Collection by hand was performed and included addition of wet napkins to clear plastic plates which housed the termites in between trials. The track was constituted of kimwipes and included ink-penned marks along pheromone-laced trails which the termites followed; this was replaced with every trial of multiple termites. Termites that refused to follow the track or did not conclude the track were removed in a short period and their results were counted as incomplete and the rates of completion will be shown in Table 2.

A stopwatch or clock was used to observe the time for a fresh randomly assigned termite to complete each track, repeated 15-20 times, and basic calculations performed to obtain the rate of travel at multiple points 3”, 6”, and 8”, which will be provided in Figure 1 and Table 1 as well as statistical significance across results to eliminate the possibility of wide variance in the research. Times were collected at each of the three points in one trial and then repeated in three more trials with each individual distance. Termites were collected after individually completing their trials and transferred through a waiting cell with a wet kimwipe back into their natural habitat, the woody environment. This provides the first set of data necessary to confirm the hypothesis: termite mid-trial rates of travel. Standard temperature and pressure was expected as well as atmospheric qualities, and the track was encapsulated by clear plastic to help ensure this. The experiment was repeated using only 3”, 6” and 8” tracks, this time with 15 trials, to help ensure data accuracy, and to confirm the hypothesis.


No instance of significant deviation was observed after statistical analysis using a p-value limit p<.05 to determine significance across multiple sets, except in comparison of different length tracks, which can be seen in Table 1 and Figure 1 below, though variance was likely too great to determine statistical significance in comparison of closer distanced tracks. Completion rates degraded as expected across the various length tracks of the termites, seen in Table 2. This establishes the premises upon which the hypothesis can be tested, as the termites were behaving in a logical fashion, and not attempting these tracks in a random sense, so statistical analysis across data sets can now be warranted.

Results were analyzed by comparing across two data sets of the same length, one three inch track mid-trial, and the other only 3 inches in sum. Comparison of 3-inch tracks with 3-inch mid-trial times results yielded a p-value of .447, indicating a statistically non-significant difference between rate of travel for termites mid-trial in comparison to a short trial, which rejects the hypothesis. Original inclusion of 2-inch tracks, raw data for which can be derived from the supplemental materials, into the Figure 1 below, did not demonstrate a significant difference in rate of travel for the late trial period, but did confirm significant difference in raw times compared to other lengths of travel, excluding 3-inch mid-trial results which is done because the p-value of .227 is likely due to variance, not in inherently significant changes in rate of travel that would result in an insignificant difference between overall times over different lengths.

Table 1: P-values between data sets obtained

3-inch mid-trial and 3-inch from start .447
3-inch trials from start .113
6-inch trials from start .432
8-inch trials .372
3-inch trials to 6-inch trials* .0016
3-inch trials to 8-inch trials* .0049
6-inch trials to 8-inch trials* .044
2-inch mid-trial to 3-inch mid-trial .227
2-inch mid-trial to 6-inch trials* 5 x 10^-5
2-inch mid-trial to 8-inch trials* 9 x 10^-6

*Denotes statistically significant difference

Table 2: Percent completion for each dataset

Datasets 3-inch mid-trial 3-inch trials from start 6-inch trials 8-inch trials 2-inch mid-trial
% Complete 35% 45-7% 35-40% 25-30% 30%

Figure 1: Average Times and Standard Deviations of Termites*

*No instances of statistical significance in comparison to mid-trial rates of travel.

Conclusion and Discussion

The greatest observations of note in this experiment are the degrading rates of completion by the termites with increasing distances, seen in Table 2, and the unwavering rate of travel in those that complete their tracks. This indicates that termites do give up traveling along pheromone laced lines when tired or distracted, but will not slow down. It is important to remove the outliers because it became apparent after preliminary observations that some termites do not have the capability of completing a track or identifying the pheromone trail which will determine their direction of travel. Some assumptions will remain in this trial, such as the amount of pheromone used, but the removal of outliers should have established a steady point to observe the individual insects, which presents in the statistical significance seen in Table 1 between distances traveled by the termites.

A significant difference between rates of travel from the start and rates of travel recorded in the middle of a trial would confirm the hypothesis that termites either tire or warm-up to complete a track, and change their rate of travel mid-trial. Failure to find a statistically significant difference between the rate of travel along three inch tracks when beginning a trial and mid-trial, seen in Figure 1, shows definitively the hardy behavior of the termites even with traveling periods of significant distances. That is a conclusion which is backed by the expression of a significant difference between trials of different lengths, necessary to show that there is not simply a wide variation in all termite travel which would indicate a necessity for retrial with a greater sample pool, or more precise laboratory equipment. For the purposes of this research, this experiment can be said to have to have rejected the hypothesis presented.

Indications of the nature of the termite behavior can be interpreted accordingly, and with cross-reference on dietary habits some insight can be drawn. Original interpretation demonstrated these percentages and differences as indicative of the termites’ stage in life, further investigation revealed this to be a function of the type of termite being used. This was confirmed in further reading among the metabolic nature of termites including an experiment from 1925 which demonstrates the consistent nature of the creatures even under such miserable conditions as acid and ash (Cleveland, 291-2).

The variation in completion of tracks proceeds logically, though is not subject to a statistical analysis, and termites were less likely to complete longer tracks than shorter tracks, as shown in Table 2, ranging from 47% at 3 inches to as low as 25% by 8 inches. Mid-trial results and 6 inch trials fall in between these results appropriately and consistently. These are the only indication of a tiring mechanism supported by data in this experiment, which warrants further investigation and is discussed to some extent with ants (Loreto, et al.) which follow bridges and perhaps scents based on absorption patterns but could be more difficult to evaluate in terms of termite behaviors.

One experiment coming to mind would be utilization of a plastic surface instead of a paper surface and perhaps a fabric one as well in order to firmly establish the impact of pheromone quantity on the results, a possible source of error discussed previously. It is worth mentioning that repetition with a greater data pool would likely establish statistical significance between 2 and 3-inch tracks, but would be unlikely to yield any change to analysis of mid-trial data, based on comparison of longer time difference analysis. This discrepancy is worth further investigation before warranting any full repetition of the data set. Another extension of this work would be to draw on the work of L. R. Cleveland and evaluate whether the longevity of termites is reflected in consistency of their rate of travel, though adequate conclusions may be possible simply from evaluation of the raw data provided here and in subsequent work to that study on the topic of termite longevity.


Cleveland, L. R. (1925) The ability of termites to live perhaps indefinitely on a diet of pure cellulose. Biological Bulletin 48, 289–293.

Loreto, R. G., Hart, A. G., Pereira, T. M., Freitas, M. L., Hughes, D. P., & Elliot, S. L. (2013). Foraging ants trade off further for faster: use of natural bridges and trunk trail permanency in carpenter ants. Naturwissenschaften, 100(10), 957-963.

Supplemental Data (in seconds):

Experiment 1: 3″, 6″, 8″ Column1 Column2 Column3 Column4
trial Start to A A to B B to C A to C
1 2 5 3 10
2 9 7 5 22
6 6 6 4
10 7
11 10 7 11 28
13 4 8 7 19
14 9 7 4 20
15 7
18 13 8

Experiment 2: 3” Experiment 3: 8″ Experiment 4: 6″
1 6 18
2 14
5 7
7 8
9 7 15
10 29
11 8
12 5 19
13 4 20
14 27 15
15 4 16 14
Skip to toolbar