William P. MacKay.
A Comparison of the Nest Phenologies of Three Species of Pogonomyrmex Harvester Ants (Hymenoptera: Formicidae).
Psyche 88:25-74, 1981.

Full text (searchable PDF, 3840K)
Durable link: http://psyche.entclub.org/88/88-025.html

The following unprocessed text is extracted from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

A COMPARISON OF THE NEST PHENOLOGIES OF
THREE SPECIES OF POGONOMYRMEX HARVESTER
ANTS (HYMENOPTERA: FORMICIDAE)*
BY WILLIAM P. MACKAY
Departamento de Entomologia
Colegio de Graduados
Escuela Superior de Agricultura
Ciudad Juarez, Chih. Mexico
INTRODUCTION
Ants are among the most abundant animals in most habitats (Petal 1967) and may even be the dominant insects in many ecosystems (Nielsen 1972; Nielsen and Jensen 1975). Harvester ants of the genus Pogonomyrmex are a major component of the energy flux through ecosystems (Golley and Gentry 1964). Ants of this genus have become increasingly important in ecological studies, including mutualism (O'Dowd and Hay 1980), competition (Mares and Rosenzweig 1978; Reichman 1979; Davidson 1980), predation (Whitford and Bryant 1979), foraging (Whitford and Ettershank 1975; Holldobler 1976a; Whitford 1976, 1978a; Davidson 1977a, b; Taylor 1977), community structure (Davidson 1977a, b; Whitford 1978b), and impact on ecosystems (Clark and Comanor 1975; Reichman 1979). It is difficult to investigate harvester ants as seasonal processes occurring inside the nest are generally unknown and the nest populations are usually underestimated. This investigation compares the nest phenologies of three species of Pogonomyrmex harvester ants: P. montanus MacKay, P. subnitidus Emery, and P. rugosus-Emery, which occur at high, mid, and low altitudes respectively. These data form the basis for a comparison of the ecological energetics of the three species (MacKay 1981).
MATERIALS AND METHODS
The species investigated.
The altitudinal comparison is based on three species of harvester This research constitutes Chapter 3 of a dissertaion submitted to the faculty of the University of California, Riverside, in partial fulfillment of the requirements for the Degree of Ph.D. in Population Biology.
Manuscript received by the editor May 28, 1981. Pu&f 88:25-74 (198 1). hup Ytpsychu einclub orglSW88-025 html



================================================================================

26 Psyche [vol. 88
ants: Pogonomyrmex montanus MacKay, P. subnitidus Emery , and P. rugosus Emery. All three belong to the subgenus Pogonomymex. Pogonomyrmex subnitidus and P. montanus are very cl osely related, both belong to the occidentalis complex (MacKay 1980). Pogonomyrmex rugosus belongs to the barbatus complex ( Cole 1968). Pogonomyrmex montanus is unusual for the genus in being a high mountain species occurring in pine forests in the mountains of southern California. Pogonomyrmex subnitidus is a mid-alt i t ude species in the San Jacinto Mountains. Pogonomyrmex subnitidus is distributed throughout southern and south central California and Baja California, occurring at lower elevations throughout much of its range. Pogonomyrmex subnitidus is sympatric with P. rugoszys in parts of Riverside County, but is uncommon in such areas. Pogonomyrmex rugosus is a low altitude species near Riverside and occurs at lower elevations throughout much of southwestern United States. It rarely occurs at higher elevations. For example, in the Joshua Tree National Monument it is present up to 1350 meters, in New Mexico it occurs at over 2100 meters. Study areas.
Populations of all three species were studied in southern Cali- fornia: P. montanus-in a yellow pine forest community between Fawnskin and Big Pine Flat at 2100 meters elevation in the San Bernardino Mountains of San Bernardino Co., P. subnitidus -in the chaparral near the Vista Grande Ranger Station at 1500 meters in the San Jacinto Mountains of Riverside Co., P. rugosus-in the coastal sage scrub community at Box Springs at 300 meters near Riverside, Riverside Co. The three species occur in clearings within these different plant communities.
Estimation of nest populations.
Two primary methods are used in the estimation of ant nest
populations: mark-recapture methods and nest excavation. Mark- recapture methods are used to compare a population before and after seasonal production. This method has been criticized as one of the assumptions is that workers mix randomly in the nest. The workers of all three species are stratified within the nests and there is strong evidence that other species are stratified as well (MacKay 1981). Also I could find no reliable way to mark the individuals such that the marks were permanent, could not be passed on to other individuals, and would not disrupt normal activities. In any case,



================================================================================

198 11 MacKay- Nest Phenologies of Pogonomyrmex 27 such a method would only estimate the numbers of foragers in a Pogonomyrmex nest, not the actual nest population. In addition, mark-recapture methods do not provide an estimate of the repro- ductive~ produced in a nest.
Excavation of nests destroys them for further study and requires a large expenditure of time and effort. I chose periodic nest excava- tion as the method of estimating production as counts of the sexuals, brood, and workers can be made. Our experience indicates that most of the nest population is collected. Pogonomyrmex spp. colonies may live 15 to 20 years (Barnes and Nearney 1953), and will live at least two years after the removal of the queen (pers. obs.). Nest longevity is unknown in the three species investigated, but based on data from other species, I expect at least 5%-10% of the nests should not have queens. The high proportion of nest queens collected (84% in P. montanus, 77% in P. subnitidus, and 80% in P. rugosus) supports the hypothesis that most of the nest population is collected. The queens do not reside in any special "queen chamber" and are of a similar size as a worker. Therefore, it is not any easier to find the queen than it is to find any individual worker in the nest. In all cases excavation was continued at least 50 cm deeper than the position of the last ant found or the end of a burrow.
Nest excavation procedure.
The procedure was as follows: The surface dimensions of the nest were determined by removal of the top 10 cm of the nest. The hole was then extended at least 50 cm on all sides. A square ditch was dug around the perimeter of the nest to a depth of one meter in the case of P. montanus nests and over 1.5 meters around the nests of P. rugosus and P. subnitidus. We were able to sit in the ditches while carefully excavating the nests in 10 cm levels. As the hole became deeper, the ditches were proportionally deepened. All of the contents of the burrows, including ants, brood, guests, stored seeds, and dirt were placed in labeled half or one liter plastic containers. Later the animals were separated from the dirt, and counted. Nest excavation usually began between 06:OO and 07:00, before the ants became active. If foragers were needed for other investigations, excavation began later in the morning or early in the afternoon. Excavation and counting of a P. montanus nest requires 6- 10 hours, of a P. subnitidus nest 20-30 hours and of a P. rugosus nest 60-90



================================================================================

28 Psyche [V -
hours. Whenever excavation was stopped to be continued on following day, the nest was covered with a heavy vinyl cloth an - cm deep layer of dirt. This was necessary to keep the inhabit; especially the males, in the nest. A total of 80 P. montanus, 2 - subnitidus, and 20 P. rugosus nests were completely excav- between 1977 and 1980.
It appeared that the excavation procedure disrupted stratifica - of individuals within the nest only slightly. When nest cham V were exposed, many individuals emerged, but most of the POT tion remained in the chambers, and assumed a defensive posg -
involving opening of the mandibles and forward extension 0 7 antennae.
The numbers of workers at each level and the position of queen were recorded. When the nests were in production, presence or absence of eggs was noted, but the eggs were counted, as they were extremely small and are easily lost in the The larvae, pupae, females, males, and callows (immature, un - pigmented workers) were counted when they were present in nests. The contents of each level were summed to obtain an e s t i i of the entire nest population.
Seed storage in nests.
The seeds were separated from the soil by filling a 1000 ml be; about % full of soil and seeds. The contents were washed into a s = with 0.5 mm mesh. The washing and swirling were continued all of the seeds were removed from the soil. The material c a u g l the sieve was washed again until only seeds remained in the s L - The seeds were then dried (60' C) to constant weight. Nest structure.
In the process of nest excavation it was noted that the g e m -
form and shape of the nests were comparable in all three spe- The P. montanus nest structure was studied by pouring a -
solution of plaster of Paris (3 tablespoons/liter of water) into nest. The solution was dilute enough that the walls of most OF tunnel system were coated with plaster. The nest was excavate- - 1-2 cm layers and the tunnel structure at each layer was m e a s i and sketched. The resulting series of "cross sections" of the - resulted in a composite drawing of the nest. Nest temperature and humidity.
Temperature data were recorded from approximately we- -



================================================================================

198 11 MacKay- Nest Phenologies of Pogonomyrmex 29 readings of thermisters permanently implanted in nests of the three species. The data were supplemented with readings taken during nest excavation, following the procedure of Rogers et al. (1972). Soil temperatures taken within the excavation hole (at least 20 cm distant from ant burrows) and within the adjacent undisturbed soil at the same level were not significantly different in two cases involving P. montanus nests (F = 0.0000 1 ns, F = 0.13ns). Similar comparisons were not made in the cases of P. rugosus and P. subnitidus as the soils were too compacted to allow the insertion of a thermometer in undisturbed soil to a depth of 30 or 40 cm. Soil samples (160 grams) were collected at various depths and oven dried (60å C) to constant weight to determine water content. At least three replicates of soil temperature and soil moisture content were collected at each level. It was anticipated that these parameters would determine the position of the brood within the nest. I assumed a correlation existed between the humidity within the burrows and water content of the soil as well as a uniformity of the soil structure in the first 100 cm of the nest where most of the seasonal changes in the positions of the inhabitants occurred. Sandy soils would release more water vapor to burrows than would clay soils, if both had the same level of soil moisture (Marshall and Holmes 1979). The amount of water present within the soil changes continuously under field conditions (Marshall 1959), which would also modify the relative humidity.
Food input into nest.
Food input was estimated by channeling the flow of foragers and sampling a fraction of foragers at regular intervals to determine the numbers of trips made and the amount of food brought back to the nest.
Twenty-eight nests of the three harvester ant species (13 P. montanus, 10 P. subnitidus, and 5 P. rugosus), were surrounded by strips of 25 gauge sheet metal. The diameters of the enclosures were approximately one meter for P. montanus, 1.5 meters for P. subnitidus, and 2 meters for P. rugosus. The sheet metal strips were buried to a depth such that 10 cm of the metal were exposed. Sheet metal with a total width of 20 cm was sufficient. The ants could not normally climb over the enclosure as the sheet metal was very smooth. The ants would occasionally begin to climb the enclosure at the junction of the two ends. In such cases the area was covered with Tanglefoot(R).




================================================================================

3 0 Psyche [VO~. 88
In some cases, especially with P. montanus, the ants would attempt to tunnel under the enclosure. When this occurred, the ants were removed from the site of the tunneling and placed near the nest entrance inside the enclosure. In such cases the tunneling was completely controlled by destroying the tunnel system and replacing it with soil.
The ants were allowed to enter and exit the colony through two 2 cm diameter vinyl tubes, 6 cm in length. Entrance of the ants to the colony through the "exit" tube was prevented by having a 0.5- 1 cm distance between the end of the tube and the soil. In a similar manner exit via the "entrance" tube was prevented. The ants were- apparently not affected by this short distance, they either simply dropped with no hesitation or rapidly climbed down from the tube to the soil. The tubes were within 15 cm of each other and were placed on the side of the nest where most of the foraging occurred. A 0.448 liter glass jar could be placed under the tube by which the ants entered the nest, thus collecting the foragers with the food items they carried. The foragers were counted and the food items collected. The foragers were released into the nest enclosure with a quantity of food (native seeds) which approximated the amount of food removed. The nests were sampled at approximately weekly intervals throughout the foraging seasons, during 1978 to 1980. All of the foragers entering P. montanus nests were collected, 1 / 5 to 1 / 6 of those entering P. subnitidus nests, and 1/60 of those entering the P. rugosus nests. With these proportions, one person could handle the activity of 5 nests during a single day. The forager populations were estimated by capturing all of the foragers throughout the day, as they returned to the nests.
Statistical analysis.
Unless otherwise indicated, the 5% level of significance was used in all comparisons. A single asterisk indicates statistical significance at the 5% level, double asterisks at the 1% level, triple asterisks indicate significance at the 0.1% level. Means are listed plus or minus one standard error. The percentages of the nest populations were used to make comparisons between the species possible. The data obtained were fit to least squares polynomial regressions (Snedecor and Cochran 1967). The curves were constructed from the equations.




================================================================================

198 11
MacKay- Nest Phenologies of Pogonomyrmex CENTIMETERS
Figure 1.
The structure of a typical Pogonomyrmex monianus nest.



================================================================================

Psyche
[Vol. 88
Nest structure.
The nest of P. montanus has numerous burrows in the upper levels (Figure 1). Below this, there is often only a single main tunnel to the bottom of the nest. Most of the ants are found in the burrows which branch from the main tunnel. The main tunnel contains few ants and is apparently used only for movement between the side burrows. In many cases there are two separate "major tunnels", as is shown in Figure 1. In P. subnitidus the two major tunnels may be separated by more than 100 cm and may appear as two separate nests. One major tunnel may contain no brood and the other may contain all of the brood in the nest. The queen and brood are usually found in the major tunnel which goes to the deeper level. The structure of the nests of P. subnitidus and P. rugosus are not shown, but are similar except that they are larger and deeper, often extending to 300 or 400 cm deep. There was no relationship between the worker populations and the nest depth (for P. montanus r = 0.1611s (65), for P. subnitidus r = 0.03ns (26), and for P. rugosus r = 0.32ns (20)).
Nest microclimatology: temperature.
The seasonal changes in nest temperatures are similar for all three species (Figure 2). The nest warms rapidly in the spring and temperatures reach a maximum at the end of June or July. The soil temperature begins to drop in August and levels out during the winter months. As the species occur at different altitudes, the temperature ranges are different. The range of P. montanus extends from slightly below zero to 20å C, that of P. subnitidus from slightly above zero to 25' C, and that of P. rugosus from slightly below 10 to 30' C.
Only the changes at the 20 and 50 cm depths are shown in Figure 2 as the other levels are similar. The differences between the levels deeper than 40 cm were generally not significant. The only important difference between the curves of the 20 cm level and 50 cm level is that the shallow level warmed sooner in the spring and cooled sooner in the fall.
Nest microclimatology: humidities.
The seasonal changes in soil moisture are similar in the nests of all three species (Figure 3). Soil moistures are high in the winter and



================================================================================

198 11 MacKay- Nest Phenologies of Pogonomyrmex 33 30-
P. subnitidus
J F M A M J J A S O N D
MONTHS
Figure 2.
Seasonal changes in the mean daily nest temperatures of three species of Pogonomyrmex harvester ants.




================================================================================

Psyche [Vol. 88
151 P montanus
15
-----
Surface
10 cm
. --.
30 cm
-
- 50cm
8
-
P montanus
--
-----
Surface
10 cm
t ---- 30 cm
, - 50 cm
f ruaosus
15
-----
Surface
10 cm
---.
30 cm
,
-
\'.
, \
- 50cm 8
---
-
100 cm = 10
"Y
c
Figure 3.
Seasonal changes in the nest humidities of three species of Pogonomyrmex harvester ants.
spring and low in the summer and fall. Throughout the winter, the soils receive relatively large amounts of rain or snow which raise the soil moistures to high levels. After this time, the surface and upper levels lose water rapidly by evaporation. The lower levels of the nest retain water throughout the entire season, although the percentage decreases. Soil moistures at levels below 30 cm are essentially the same for all three species. Summer showers rapidly increase soil moistures of the upper levels (note the peaks in the Figure 3), but have little effect on the levels below 30 cm. This water input into the soil is rapidly lost by evaporation.
The soil moisture of the lower levels is generally higher than that of the upper levels, possibly forming a relative humidity gradient. There are more fluctuations in the higher levels, both in soil



================================================================================

198 11 MacKay- Nest Phenologies of Pogonomyrmex 35 moisture and temperature. This probably accounts for much of the brood being kept in the lower nests levels. The harvester ants apparently obtain water from several sources. Some metabolic water may be available to the ants, as it has been shown that harvester ants increase their metabolism when they are water stressed, without increasing their activity (Ettershank and Whitford 1973; Kay and Whitford 1975). Morning dew would not normally be available as foraging begins after dew has evaporated. I have seen harvester ants actively drink rain drops on the soil surface, demonstrating a curious pumping action of the gaster, but precipitation is not common in the three habitats during the summer (U.S. Weather Bureau Climatological Data). Capillary condensa- tion occurs in the soil at relative humidities above eighty percent (Rode 1955) and may allow the ants free water. Arthropods, especially insects, are able to actively absorb water vapor from unsaturated air, although the mechanism is not understood (Edney 1974; Cloudsley-Thompson 1975). It is not known if harvester ants have the ability to actively absorb water vapor. Seasonal changes in nest populations.
The data on nest populations obtained from the nest excavations are summarized in Appendix 1. Absolute counts could not be easily compared because the numbers of individuals present in the nests of the three species are very different. To reduce this variation between nest populations of the three species, the data are compared in the form of percentages. The seasonal changes in the brood and sexual populations are similar for all three species, when the percentage composition of each of the classes are compared (Figs. 4 & 5). In the three species, egg laying begins in late April to late May, similar to P. owyheei (Willard and Crowell 1965) and P. occidentalis (Lavigne 1969). Development from egg to callow in the species requires five to six weeks compared to 25 days for P. badius (Gentry 1974) and 30 days in P. occidentals (Cole 1934). It is very difficult to determine the number of larval instars in the development of ants (Wheeler and Wheeler l976), although Marcus (1 953) suggests that there are four instars in P. marcusi. As a consequence, all of the instars were combined into a single group. The first larvae appear about a week after the eggs are laid, first pupae about two weeks later. Callows are found in the nest about 5 or 6 weeks after the eggs were laid and



================================================================================

Psyche
[Vol. 88
301 P. rnontanus o
0 B0
' . Larvae 1978
.... - ..
( Larvae I979
Pupoe 1978
A ( PUPW 1979
å Callows 1978
. C- ( Caiiows 1979
L
J F M A
M J J A S O
MONTHS
b
J F M A
0 ......... Larvae
A- Pupoe
0--- C0llow
-
J F M
Figure 4.
Seasonal changes in the brood populations of three species of Pogonom.vrmex harvester ants. The arrows indicate the dates when eggs were first found in the nests. Nests excavated which contained only adult workers are not represented in the figure.




================================================================================

I 98 I] MacKay- Nest Phenologies of Pogonomyrmex 37 remain pale for about three weeks. Thus, development from the egg through the larval instars requires about three weeks, the pupal stage 2-3 weeks, and the callow stage three weeks. Most of the eggs are laid in the spring as large amounts are found early in the season. The amounts found in later excavations decrease