Environmental Engineering of Pogonomyrmex Harvester Ant Mounds

Part of the Young Naturalist Awards Curriculum Collection.

by Anna, Grade 11, New Mexico - 2006 YNA Winner

Today, the human population is growing beyond our planet's ability to support it. Groans from Americans as they grudgingly fork out their savings to pay exorbitant gas prices reveals that an alternate source of energy is desperately needed, not only to maintain the lifestyle of Western civilization, but also to stop further global warming, the consequences of which are unpredictable and potentially catastrophic. For a while now, the idea of harnessing the immense power of the sun has been investigated with some degree of success but has never really been achieved in a practical and cost-effective manner. Yet in other, more "primitive" societies, this feat may have already been accomplished. With this in mind, I have examined the habits of some of our fellow creatures in an attempt to find more effective techniques for harvesting solar power.

Dry scrubby land with sparse bunches of grass, low growing trees, and a grayish-colored anthill.
Anthill #1 in White Rock Canyon

On a crisp morning, soon to turn into a scorching summer day, I walk along a trail leading to White Rock Canyon, near Los Alamos, New Mexico. Many would not consider this sparse woodland habitat, which is covered in scrub oak, piñon, and juniper, to be attractive. However, I grew up here, and although this harsh wilderness may not possess the majesty of a towering pine forest, or the awe-inspiring splendor of a rain forest, to me it has its own stark, rugged beauty. Along this path it would be hard to miss the towering anthills built by Pogonomyrmex, the harvester ant. At first glance their mounds seem to be composed of small jewels, but upon closer examination they are actually built out of tiny quartz crystals about 1-2 mm in length. For a long time I considered these shimmering palaces to be just for show, true "diamonds in the rough" dotting the desert landscape. But eventually I began to speculate on the purpose of these quartz mounds (Dinger, 2001). Perhaps these tiny creatures use the light-transmitting properties of quartz to trap heat, which is then absorbed in the mounds, in order to warm their habitats. More explicitly, because quartz is transparent to visible and ultraviolet light from the sun, the sunlight is absorbed, heating the mound. The layer of quartz on the anthill then acts as insulation, preventing heat from escaping. Since ants are most active in warm temperatures, such heat retention inside the mound could increase the time available for ant activity.

Photo of an ant hill on rocky patch of ground surrounded by shrubs.
Anthill #2 in White Rock Canyon

The longer I considered this intriguing hypothesis, the more determined I became to understand the function of the quartz crystals glittering atop the harvester ants' mounds. Perhaps this investigation of these miniscule fascinations will inspire similar ideas for our society, enabling the solution of mankind's energy and environmental problems.



Ant colonies are both fascinating and complex (Gordon, 1999), although much about the lives of the harvester ant and other species is still unknown. A colony begins with the mating of the tiny winged alates. The females become potential queens after mating; however, their chance of survival is less than 1 percent. The males die within a few days of the mating ritual. Against the odds, a few novice queens manage to create colonies, which become established only after two years. After surviving this crucial era, colonies go through various phases, each with a unique colonial personality, until they reach six years of age, at which time the population growth plateaus and the colony remains at a size of about 10,000 ants for 12 more years. The queen sustains the colony, although her job is far from royal. Her only function is to produce larvae for her colony until she dies, after 20 years. Once the queen is dead, the colony will also die out, since its population can no longer be replenished.

The colony inside the mound consists of a complex array of tunnels and small chambers reaching about three meters into the ground. Some of this immense space is used for food storage, while other chambers house the ant larvae. The behavior of the ants within the colony mirrors their home's complex structure. Such a society has strangely communistic attributes. Each ant attends to its own duty with no apparent command structure; all the ants work together in harmony for the good of the colony. Communication between the ants is primarily thought to be through scent and by touching antennae. All ants are in their adult stage when they break out of their pupae. A newborn ant tends to the pupae and larvae. As an individual ant grows older, it moves towards the surface of the colony: the next task in the life of an ant is to stack seeds. Finally, an ant moves to the outside to work as a forager, nest maintenance worker, patroller, mitten worker (trash collector), or reservist, performing whatever task is needed (Folgarait, 1998).

Photo of ants at entrance to their ant hill.
Ants at the entrance to their mound.

Overall, outside workers make up about 25 percent of the ant population in a colony and only live at the top of the nest for the last two out of the 12 months of their lives (Holldobler, 1990, 1994). In the above photo, ants are seen outside the entrance of their mound, actively foraging. The food the ants gather is mainly used to feed the larval population, not the active ants. Far more crucial than food to harvester ants is moisture. None of the harvester ants drink; they depend on their surroundings to be wet enough to sustain them, so it is crucial that their mounds retain moisture in the scorchingly dry conditions in which they live (Nkem, 2000). All of these characteristics may drive harvester ants to engineer their mounds to preserve heat and moisture.



Hand-drawn map of a trail into a canyon.
Figure 1: Map of the trail heading into White Rock Canyon

I began my study of the  Pogonomyrmex  harvester ants (Werner, 1994) by scouting out four ant nests on a roughly one-kilometer stretch of the trail heading into White Rock Canyon, shown in the map in Figure 1.

The size of the mounds varied from a diameter of approximately 25 inches and a height of four inches for Mounds 3 and 4, to a diameter of 33 inches and height of 7 inches for Mound 1, to a diameter of 50 inches and height of 7 inches for Mound 2. I observed the ants' behavior, their habitat, and their nests during the summer of 2005 to form the basis for my project. What I discovered was that these ant nests are not just covered with quartz crystals, but are composed of them. By shifting through a small sample of a mound, I determined that the mounds are a mixture of quartz and fine dirt, in a 2:1 ratio by volume, with the dirt hardening into a cement-like mortar to bind the mixture together.

From this initial study, I formulated the three different phases of my experiment. First, I spent 12 hours, from 6 a.m. to 6 p.m., testing the temperature inside the four anthills in the sketch, the ground next to them, the air temperature, and the surface temperature, in roughly one-hour increments, while simultaneously observing the ants' activity (the number of ants outside the mound and distance they traveled away from the mound). The number of ants was estimated by roughly counting the number of ants on the mound at a particular time, backed up by a photo of the mound taken at that time. To determine the distance the ants traveled, I measured the distance from the mound to the farthest ant I saw. These measurements served to reveal if and how ant activity correlates with temperature, as well as whether the anthills do retain heat better with quartz than simply with dirt.

Next, I obtained pieces of various rock samples that were approximately 1-2 millimeters long: turquoise, rose quartz, peridot, citrine, blue topaz, quartz, and a sample of mixed gemstones, all with different absorption spectra, and placed them next to Mound 2, one of the two large anthills. Over the next three weeks I observed the use of these stones by the ants.

Finally, I built two anthills, without ants, so I could do more extensive temperature tests with no further disturbance of the ants. One hill contained the precise quartz mixture of the anthills I had observed, 33 percent fine soil and 67 percent quartz pebbles, while the other was made up of dirt collected from the area surrounding the anthills studied. Both hills were the same volume. I observed the temperature of these anthills, in one-hour increments, over two 24-hour periods.

The concluding element of my project I stumbled upon during the study of the temperatures in my man-made anthills. I had placed plastic bags over the anthills so they would not get wet when it rained. It had been nearly a week since the last summer shower, and as I removed the plastic bag from the quartz anthill I observed condensation on the inside of the bag, whereas the solid dirt anthill revealed no condensation. I decided to investigate whether the quartz anthill retained moisture better than the dirt anthill by mixing an identical amount of water (0.25 liters) into the two hills and then weighing them twice a day over three days, taking a loss in weight as a loss in moisture. I repeated this measurement four times and averaged the data.

Figures 2 and 3  (Click to enlarge)
Figure 2: Air, mound, ground and surface temperatures at anthill 1 (Click to enlarge)
Chart of surface temperature against the number of ants outside an anthill.
Figure 3: Surface temperature versus the number of ants outside for anthill 1 (Click to enlarge)
Chart of activity over time for anthills.
Figure 4: Ant activity over time for all anthills. (Click to enlarge)

Results and Analysis

First, I will describe the results of the field study of the four anthills. Figure 2 reveals the temperatures (air, mound, ground, and surface) over time at Mound 1. The surface temperature exhibits a large variation and responds quickly to changes in the environment, such as the clouds that appeared from noon to 2 p.m. The other temperatures (air, ground, and anthill) increase slowly and steadily as the day heats up. In Figure 3, I plot the surface temperature against the number of ants outside for Mound 1. As is seen in Figure 3, the number of ants outside the anthill correlates most readily with surface temperature, with ant activity maximizing at approximately 90° to 100°F., but plummeting to zero above 115°F. (Note that no ants are active in the dark.) Finally, I measured the distance that the ants traveled from the entrance to the mound as a function of time. This data, plotted on the same graph as the surface temperature and the number of ants on the mound, is shown in Figure 3. It reveals that the distance that the ants travel peaks at about the same temperature as the number of ants and so also correlates with the surface temperature. This result is reasonable because one would expect the ants to forage farther under optimal temperature conditions when there are more ants with which to compete.

Figure 4 reveals the ant activity for all four of the mounds. All exhibit the same qualitative behavior, maximizing their activity at the same time of day, either during cooler temperatures in general, or during episodes of cloud cover that allow the surface temperature to drop. Different anthills display different maximum numbers of ants, since these anthills were various sizes. The ants were quite sensitive to temperature, and when a cloud passed in front of the sun, which almost immediately cooled the surface temperature, the ants responded immediately, leaving the mound to forage.

Chart of total number of ants compared to surface temperature.
Figure 5: Surface temperature and total number of ants from four anthills over time (Click to enlarge)
Chart of number of ants against surface temperature.
Figure 6: Total number of ants compared to surface temperature (Click to enlarge)

This response is seen clearly in Figures 5 and 6, where I plot the total number of ants from all four mounds and the surface temperature on the same graph (Figure 5), and the number of ants compared to surface temperature (Figure 6). I made plots of the number of ants versus all of the other temperatures measured (ground, air, and anthill), but the only temperature on which the number of ants depended was the surface temperature. Again, this result is reasonable because the ants scampered along the surface as they carried out their tasks outside the mounds.

This study was carried out during the summer, so it became unbearably hot during the day. The data reveals that for certain periods of the day the heat was actually a problem for the ants, too, essentially limiting their activity. Why would the ants want their hills heated if at some times of the day they could not bear to step outside for fear of being fried? This puzzle definitely provided food for thought.

Photo of rocks placed around an ant hill.
Photo of rocks placed around an ant hill.
Original rocks placed on Mound 2 (top) Remaining rocks 48 hours later (bottom)

The next phase of my project, observing the ants transporting crystals placed near a mound, revealed that the ants did move the crystals. Many crystals were observed on top of the mound adjacent to the gemstones, and some gemstones were observed on a neighboring mound 50 feet away. The photo (top) shows the crystals just after they were placed by Mound 2, while the photo (bottom) reveals the smaller remainder after 48 hours. Of the various small crystals I placed by Mound 2, the most popular with the ants was turquoise, a rock that does not transmit light, thus one that would heat up the anthills the least. I suspect that this stone was the most popular because the flakes of turquoise were slightly smaller than the other stones, making them more convenient to carry.

Photo of model of ant mound outdoors.
Figure 8: model mound using only dirt surrounding the ants' habitat

Building my own anthills to test their temperature became crucial so that I could determine if quartz retained the heat better overnight, a time when the ants might actually want warmth. I constructed a model ant mound using the measured ratio (2:1) of quartz to dirt, and a second model mound, shown in Figure 8, constructed only of soil from the same area. The mounds were both 2.5 inches high and 6 inches in diameter. Figure 9 and Figure 10 show the temperature data, taken on two different days. I recorded the temperatures of the surface (green), the air in the shade (blue), in the deep ground (pink), as well inside the quartz (black) and dirt (red) mounds. The presence of quartz rocks increased the temperature of the mound by an average of 2.5°F., plus or minus 0.5°F., compared to the mound without quartz, a small but consistent effect that might be especially important at night. A similar increase was seen in both sets of data, although the air temperatures on the two days were different. Furthermore, the increase in temperature was consistent throughout the night, when such an increase would be beneficial for the ants' activity level.

Figure 9 Anna
Figure 9: Temperature data for model mounds on first day (Click to enlarge)
Figure 10 Anna
Figure 10: Temperature data for model mounds on second day (Click to Enlarge)
Chart showing moisture loss for ant mound with quartz and mound without quartz
Figure 11: Moisture loss for the mound with quartz and the mound without quartz (Click to enlarge)

Feeling dejected because the measured increase in temperature in the quartz mound was rather small, I was preparing to disassemble my makeshift anthills when the condensation on the inside of the plastic bag led me to the hypothesis that perhaps anthills built with small stones retain moisture better than those built only from dirt. To determine how well each mound retained moisture, I mixed 0.25 liter of water into each and measured their weights as a function of time. I took four sets of data over a period of a couple of weeks. Figure 11 shows the average of this data, for the quartz (black) and dirt (red) mounds, revealing the superior moisture-retaining capacity of the quartz mound. The amount of water, Wi(t), declines linearly with time for both mounds: Wi(t)= 1- Kit, where kquartz  = 0.2/per day and kdirt  = 0.33/day. Therefore, moisture is retained for 66 percent longer in the quartz mound. This linear decrease in moisture over time, along with the smaller proportionality constant for the quartz mound, is consistent with a moisture-loss model in which the moisture loss increases with exposed surface area, and the quartz partially blocks (decreases) the surface area. Hence, the quartz rocks form a partial barrier preventing moisture from escaping from the mound. Since a harvester ant's need for moisture shapes its lifestyle—from how long it can stay out of its nest to its construction of extensive underground tunnels—it would be reasonable for these ants to build their nests with stones to retain moisture.



I investigated the reasons why the harvester ants near my home cover their mounds with quartz crystals, with the hypothesis that the ants use quartz to regulate the temperature and humidity within their mounds. I found that ant activity correlates with surface temperature, with maximum activity around T=95°F. and no activity for T>115°F. Ant mounds consist of small rocks to dirt in a 2:1 ratio, and ants actively carry small rocks to mounds. The presence of quartz rocks increases the temperature in the mound by an average of 2.5°F., plus or minus 0.5°F., a small but consistent effect that is perhaps most important in the cool of the night. The quartz rocks also increase the length of time that moisture is retained, by approximately 66 percent. Building anthills with small stones could have other practical uses that I have not yet investigated, such as resistance to erosion or ease of mound construction. However, it is clear that trapping moisture is an important feature of the quartz mound design, enabling the ant to maintain a viable moisture level in their habitats in this desert environment.

This project reveals that ants have successfully controlled their environment by simple quartz engineering. Perhaps mankind can imitate ants in the quest for clean solutions to the energy crisis. Moreover, these tiny creatures, which many people unwittingly crush under their shoes, inspire many questions for further exploration. I could study the physical characteristics of the anthills during seasons other than summer, because with lower temperatures the insulating effect of the quartz may be more important, and its effect on moisture retention less important. Alternatively, I could examine how ants affect their environment. For example, what are the environmental consequences of the presence of these ant mounds? Finally, I could redirect my investigations towards ants' behavior towards one another, determining how they interact with members of their own and other colonies. All of these studies are important because they reveal how other life forms build habitats to suit their own needs, and how they utilize the natural resources available to them and integrate them into their lifestyle. If one does science to satisfy one's curiosity, then eventually someone will stumble across a discovery crucial to mankind's survival. After all, many of the most important scientific discoveries through the ages were accidental.




Gordon, Deborah. Ants at Work. New York: W.W. Norton and Co., 1999.

Holldobler, Bert, and Edward Wilson. Ants. Cambridge, MA: Harvard University Press, 1990.

Holldobler, Bert, and Edward Wilson. Journey to the Ant. Cambridge, MA: Harvard University Press, 1994.

Werner, Floyd, and Carl Olson. Insects of the Southwest. New York: Da Capo Press, 1994.


Journal Articles

Dinger, Wolfram, and Manfred Wanner. "Development of soil fauna at mine sites during 46 years of reforestation." Pedobiologia 45 (2001): 243-271.

Folgarait, Patricia A. "Ant diversity and its relationship to a functioning ecosystem: a review." Biodiversity and Conservation 7 (1998): 1221-1244.

Nkem, J.N., et al. "The impact of bioturbation and foraging on surrounding soil properties." Pedobiologia 44 (2000): 609-621.