David Brown, Darren Butler, Jeremy Hawkes and Heather Marsh.
Abstract. Western Harvester ants abide primarily in the deserts west of the Mississippi River and can be found in spots of mounding soil and extensive tunneling (Gordon, 1995). The specificity of this specie s habitat location suggests that the ants have precise adaptations to certain environmental conditions such as temperature, humidity, food availability, predators, and soil type. By controlling these factors but varying soil type, it appears that the ants are best adapted to the more coarse soil. With a higher tunneling volume and rate, the ants are most efficient in coarser soil which must be one of the required elements that bring these ants to their current habitat.
Harvester ants are large ants of the pogonomyrmex species (Crist, 1999). They can be up to + inch in length and will be seen crawling on cleared areas of ground around the home. They may be seen going in and out of craters or large holes in the ground which are surrounded by cleared areas from 3 to 35 feet in diameter (Johnson, 2000). Western Harvester ants make a small mound on top, but then tunnel up to 15 feet straight down to hibernate during winter. Ant mounds consist of many chambers connected by tunnels. Different chambers are used for nurseries, food storage, and resting places for the worker ants.
The type of habitat the Harvester ants select may be based on several different variables such as climate, food availability, predators, soil type, etc (Crist, 1999). By removing all other variables but changing the soil type, the ants will demonstrate which soil coarseness they are best suited for.
Eight, Ball wide mouth jars where used to contain 15 ants each. Black posterboard was cut into strips and where stapled together to form a tube. The tubes were placed inside the wide mouth jars. In two of the jars, one of two soil types where poured between the outside of the jar and the posterboard insert. Slits were cut into the top of the jar lids and aluminum foil was wrapped around the outside of the jars. The aluminum foil didn’t cover the jar completely, only coming up to a half inch of the top of the jar. Food (eggs, flour, and honey) was then place inside the posterboard tube in the bottom of the jar. See Figure 1.
To measure tunneling the jars were checked periodically to look for visible tunnels. The tunnels were marked on the outside of the jars with permanent markers. Each time period was marked differently so that individual and composite growth could be seen. After the jars were emptied the living ants remaining were counted.
Jar 1: (sand) 11 ants remaining
Jar 2: (sand) 11 ants remaining
Jar 3: (brown soil) 10 ants remaining
Jar 4: (brown soil) 11 ants remaining
The marked empty jars were then measured. We overlaid a grid system on the markings. The area was calculated by the size and amount of squares in the tunneled through areas. The individual jars were measured and plotted on a graph to see comparative growth on a time axis.
Jar 1: Sand
After 42 hours: 11.35 cm2
After 54 hours: 21.67 cm2
After 65 hours: 31.48 cm2
After 68 hours: 34.45 cm2
Jar 2: Sand
After 42 hours: 37.16 cm2
After 54 hours: 64.26 cm2
After 65 hours: 70.97 cm2
After 68 hours: 72.00 cm2
Jar 3: Brown Dirt
After 42 hours: 16.26 cm2
After 54 hours: 38.20 cm2
After 65 hours: 44.65 cm2
After 68 hours: 44.65 cm2
Jar 4: Brown Dirt
After 42 hours: 10.84 cm2
After 54 hours: 16.26 cm2
After 65 hours: 17.29 cm2
After 68 hours: 19.87 cm2
Series 1 is jar 1, series 2 etc.
Time 1 is 42 hours, 2 is 54, 3 is 65, and 4 is 68
We then took the total amount of tunneling in each jar and divided it by the amount of ants still alive. We plotted this to compare tunneling per ant in the four jars. Tunnel Area per ant: Jar 1: 3.132, jar 2: 6.54, Jar 3: 4.47, Jar 4: 1.81s
There were 15 ants put into to watch jar originally, I checked the jars after 4, 8, and 20 with no noticeable tunneling. After 42 hours I first saw tunneling. At 42 hours I checked the native soil ants and found that they were all stuck in middle dead.
In jar 4 there wasn t much visible tunneling. However most of the area that was visible was spread out into many different sections in the bottom. Most of the tunneling that lead to these sections was not seen, so there was actually a lot more tunneling than the data shows.
Unfortunately, the most significant conclusion to which we arrive is that a wiser construction of artificial habitat is recommended when placing these ants in their native soil. It appears that the dust from the arid soil (not found in the others) coated the inside of the container and prohibited the ants from returning to the soil to dig. This limits us somewhat in the conclusions that can be drawn, as there are no conclusive results (besides the fact that the ants are conclusively dead) from two of our six habitats.
There is a significant difference, however, in the results obtained from the more coarse soil and the finer soil. With regard to both the amount of tunneling and the rate at which the tunneling occurs, under our experimental conditions, coarser soil seems to be more favorable to these harvester ants. In the case of each sample, the rate of tunneling is greatest as the ants begin their activity, but then declines over time, eventually reaching a rate near zero. The initial rates are much higher in the coarse soil (average initial slope of 11.4) than in the fine soil (avg. initial rate of 4.75).
As each sample included some dead ants, we thought it profitable to compare the amounts of digging per living ant, and again found the average of the coarse sand to be much higher than the fine dirt (4.84 compared to 3.14, see graph above).
These findings indicate that when other variables are removed, there seems to be a trend towards greater activity when the soil is more particulate and coarse. Obviously this leaves much more to be discovered and analyzed by further study, but this does lend some understanding as to why harvester ants are more abundant in the hard-packed arid regions of the country rather than the more fertile and soft soils.
Further studies would seek to identify the ancillary conditions that lead to habitat favorability; conditions such as temperature, moisture, types of ground cover, predator prevalence, and food abundance. Were we to conduct our study again, we would make greater efforts to bring our experimental conditions into greater harmony with those occurring naturally, especially where temperature (ideally maintained around at 32 degrees C, rather than the 26 that we had), native nutrition, and (especially in the case of arid soil) mobility are concerned. After observing some variation in the results among identical soil classes, it also seems prudent to increase the sample size and repetition of our methods. Works Cited