Abstract
The discovery of numerous Pyramica ohioensis and P. rostrata colonies living in acorns, as well as the efficient recovery of colonies from artificial nests placed in suitable habitats, opens a new stage in the study of North American dacetine ants. Here we present detailed information, based on 42 nest collections, on the colony structure of these two species. P. ohioensis colonies are smaller than those of P. rostrata. Both species are polygynous, but nests of P. ohioensis contain fewer dealate queens than those of P. rostrata. This is the first report of multiple collections of Pyramica colonies nesting in fallen acorns, and of the use of artificial nesting cavities to sample for dacetines in the soil and leaf litter. We describe an artificial cavity nest design that may prove useful in future investigations.
1. Introduction
The ants of the genus Pyramica are distributed worldwide and are diverse and abundant in both warm temperate and tropical forest communities (Bolton lists 350 valid species) [1]. In the United States, Pyramica is represented by 32 native species and 9 introduced species [2–5]. Two additional undescribed species from the Southwest have been discovered recently (Cover and Alpert, [6]). Pyramica species are widely known as specialist predators of Collembola [3, 7–16], however, little is known about other aspects of their biology. These diminutive, cryptobiotic ants forage in leaf litter and rotten wood and are most frequently collected by litter sifting, Berlese funnel extraction, or Winkler samples. Collections of colonies are comparatively infrequent and thus our understanding of Pyramica colony demographics and life histories is limited.
This study began as a trial of the use of artificial nests to sample soil and leaf litter ants in a Maryland oak-hickory forest. Sets of artificial nests designed to evaluate preferences for cavity volume and diameter of the nest entrance were placed at the study site in the spring and retrieved at the end of the summer. Although a variety of ant species colonized the artificial nests, the most interesting result was the recovery of several colonies of dacetine ants belonging to the genus Pyramica. Based on this finding, it was assumed that Pyramica was nesting in natural preformed cavities in the same general area where the artificial nests were placed. Pyramica ohioensis and P. rostrata were eventually discovered colonizing acorns. Once the appropriate conditions were recognized, several Pyramica colonies were consistently recovered on each collecting trip.
Pyramica ohioensis (Kennedy & Schramm) (Figure 1) can be recognized by a combination of unique characters including J-shaped hairs on the lateral clypeal margin pointing posteriorly, absence of a gap between the basal lamella and the first basal tooth. All curved hairs on the ventral margin of the scape are pointing toward the apex. State records: Alabama, Arkansas, Delaware, District of Columbia, Florida, Georgia, Illinois, Indiana, Kansas, Kentucky, Louisiana, Maryland, Missouri, New Jersey, North Carolina, Ohio, Oklahoma, Tennessee, Texas, and Virginia.
Pyramica rostrata (Emery) (Figures 2 and 6) can be recognized by a combination of characters including J-shaped hairs on the lateral clypeal margin pointing anteriorly, having a small gap between the basal lamella and the first basal tooth, and curved hairs on the ventral margin of the scape pointing toward the base. State records: Alabama, Arkansas, District of Columbia, Florida, Georgia, Illinois, Indiana, Kentucky, Louisiana, Maryland, Mississippi, Missouri, New Jersey, North Carolina, Ohio, Pennsylvania, South Carolina, Tennessee, Texas, and Virginia.
This is the first report of multiple collections of Pyramica colonies in acorns. Although hollowed-out stems have been used in other studies in the tropics, we believe that our artificial nests also represent the first application of artificial nesting cavities to sample Pyramica species. We describe a design of artificial nests that may prove useful in future investigations of species of ants which form small colonies.
2. Materials and Methods
2.1. Artificial Nest Design
Each artificial nest consisted of three pieces of pine (Figure 3), 5.0 cm × 3.4 cm × 0.5 cm. The middle insert had two holes (diameters 0.63 cm and 1.26 cm, 0.2 cm from each edge) drilled through it to create nesting cavities. Entrance holes of different diameters were drilled into each cavity from the side: 1.56 mm, 1.95 mm, and 2.35 mm. Glass cover slips were glued over the top and bottom of each cavity to facilitate inspection of the artificial nest and to retain inhabitants when the nests were retrieved and opened. The layers were tied together with monofilament fish line. Each trap was given an accession number which identified the diameter of the cavity, the diameter of the entrance hole, and the trap number.
2.2. Artificial Nest Placement
The ant nests were randomly placed at the study site in groups of five (referred to as a trap line). Traps varied by cavity volume and/or entrance diameter. Each trap was attached to a 3–5 m piece of monofilament and spaced evenly. Each trap and trap line was buried in the leaf litter.
2.3. Study Site
The artificial nests were placed in the woods associated with the Croydon Creek Nature Center, Rockville, Montgomery County, Maryland. The GPS coordinates are 39°05′23′′ N; 77°07′42′′ W. The woods are dominated by multiple species of oak, hickory, beech, maple, and tulip. Deep ravines cut through the woods.
2.4. Retrieval of Artificial Nests
Each trap was placed individually into a resealable plastic bag. The nests were refrigerated at 3°C until they were processed.
Each cavity was then inspected for the presence of ants or other organisms. If the trap cavity contained an ant colony, the entrance hole was plugged and later inspected with a dissecting microscope.
Each ant colony was placed in 70% ethyl alcohol. Much care was taken to remove all larvae, pupae, and adult ants from the nesting cavity. Ant material was placed in two dram screw cap vials with poly-seal lids. Each sample was labeled with the appropriate collection information and accession number.
2.5. Acorn Collection
After it was discovered that multiple Pyramica colonies occupied the trap nests, a concerted effort was made to find natural colonies. In the same woods where the trap lines were placed, within 0.25 of a mile, a colony of Pyramica was discovered in a decomposing acorn. As work proceeded, Pyramica colonies were recovered repeatedly from acorns.
When acorns were found on the ground, the top leaf litter was brushed aside exposing the organic horizon. Acorns that had entrance holes and that were embedded in the top of the organic soil horizon proved to be the best candidates for finding Pyramica colonies. Acorns that were in an advanced stage of decomposition were ideal. In order to locate these acorns, the top layer of soil was disrupted by running fingers over it in a claw-like motion. Many times, acorns were dislodged that were not otherwise visible. Acorns were cracked open but not split into two disconnected halves. Colonies of of Temnothorax were easy to detect. Sometimes the clumped, white larvae were seen first or frequently the workers would rush out of the acorn when disturbed. If the acorn contained a Ponera colony and if the light-tan-colored pupae were present, they were the first objects observed. Often it was possible to observe the shiny appearance of disturbed workers scurrying around. If an acorn was split open and contained a Pyramica colony, the workers remained motionless. The open acorns would have to be held in direct sunlight to see the cryptic slow moving Pyramica workers. Secondly, larvae were uniformly spread around the cavity and were very small so that they were not noticeable except under a microscope. Although the workers were cryptic, the light colored spongiform gland/tissue on the petiole and gaster [17] helped to recognize workers.
2.6. Processing Acorns
When an acorn was located that contained a Pyramica colony, it was placed in a 50 mL screw-cap plastic conical centrifuge tube. Most of the acorn samples were preserved with 70% ethyl alcohol. A few colonies were kept alive so that the ants within could be observed and photographed.
2.7. Processing Ant Samples
The acorn collections were processed by the same procedure as with the artificial ant nest samples. For each sample, total number of workers, dealate queens, male and female alates, pupae, and larvae were counted and recorded.
2.8. Deposition of Specimens
All materials were identified by the second author [G. D. Alpert]. Voucher specimens have been placed in the Museum of Comparative Zoology, Harvard University. The remaining materials reside in the collection of the first author [R. M. Duffield].
3. Results
Of the 55 artificial nests (110 nesting cavities) placed in the woods, 18 cavities (16%) contained ant colonies or ants. Five different species of ants were recovered from the artificial nests. They included Temnothorax curvispinosus (6 colonies), Temnothorax longispinosus (1 colony), Ponera pennsylvanica (3 colonies), Pyramica ohioensis (Kennedy & Schramm) (2 colonies), and P. rostrata (Emery) (6 colonies).
Two of the artificial nests had both of the nesting cavities occupied by ant colonies. One had two T. curvispinosus colonies residing in it. The second nest had a T. curvispinosus colony on one side and a Pyramica colony on the other side. Six of the eighteen cavities occupied by ant colonies were 6 mm in diameter. Pyramica colonies occupied two 6 mm diameter cavities; six others colonized 1.2 cm diameter cavities.
A total of 29 P. rostrata colonies were recovered: 6 from artificial nests and the rest from acorns. The demographic data of each colony are presented in Table 1. We did not want to mix the colony data recovered from the artificial nests with the acorn data. Since there were more samples for each species from acorns, these were the data that were analyzed.
3.1. Demographic Data for P. rostrata Colonies Recovered from Acorns
Analysis of colonies recovered from acorns provided the following data: mean number of workers 82.9 ± S.D. 39.1; range 14–160, ; mean number of queens 5.0 ± S.D. 4.8; range 0–24, ; mean number of larvae 63.3 ± S.D. 38.0; range 0–140, . Eight colonies contained over one hundred adult workers. A total of 2,074 P. rostrata workers were collected during this study. While most of the P. rostrata colonies contained multiple queens, one contained no queen and two others contained single queens. Surprisingly, one colony contained twenty-four queens. Five of the colonies contained at least one male and/or female alate. These reproductives were primarily recovered from the artificial nests. Most of the recovered P. rostrata colonies contained larvae. The greatest number of larvae recovered from a single nest was 140. Larvae appeared to be one uniform age class. Seven colonies contained at least one pupa.
3.2. Demographic Data for P. ohioensis Colonies Recovered from Acorns
The demographic data of each of the 14 Pyramica ohioensis colonies recovered from acorns is given in Table 2. Analysis of colonies provided the following data: mean number of workers 55.5 ± S.D. 23.5; range 19–111, ; mean number of queens 2.6 ± S.D. 1.7; range 1–6, ; mean number of larvae 39.4 ± S.D 18.9; range 0–83, . The number of adult individuals per colony ranges from 10 to 83. Only one colony contained over one hundred adult ants. A total of 763 workers of P. ohioensis were collected during this study. Five of the colonies contained single queens while the rest contained multiple queens. The greatest number of queens in a single colony was six. Three colonies contained at least one alate. Five of the colonies contained one or more pupae. The largest number of larvae recovered was eighty-three.
The number of workers per colony versus the number of larvae was positively correlated at 63% for P. rostrata and 67% for P. ohioensis. This may be a reflection of the number of workers required to support development of a population of larvae since recruitment is based upon individual foragers returning with Collembola as prey.
However, the number of queens did not correlate strongly with either the number of larvae (17%, 25%) or the number of workers (.07%, 17%), P. rostrata and P. ohionesis, respectively. This preliminary data suggests that multiple queens are not laying viable eggs at the same time and that there may be a single reproductive queen. All the brood was synchronized in development as overwintering larvae. This colony reproductive system needs to be explored further to understand the role of polygyny in these two species.
3.3. Overwintering Colonies
During the collecting phase of this investigation, several tubes with acorns containing Pyramica colonies were left refrigerated [3°C] for more than three months (end of October through February 1, 2011). Each colony was then checked using a microscope at low magnification. The workers showed no movement at first but were active by the next day.
In retrospect, being able to refrigerate Pyramica colonies for months and revive the specimens at room temperature is not surprising. Temperate ants must survive periods of subfreezing temperatures with minimal losses. It may prove useful to store Pyramica colonies in natural cavities for laboratory experimentation. Colonies could be retrieved from storage as needed and thus reduce the burden of keeping live laboratory colonies.
4. Discussion
4.1. Artificial Nests
Although P. rostrata and P. ohioensis have been reported in Maryland and Virginia [19, 20], this is the first report of the recovery of multiple colonies of each species in artificial nests. As Deyrup and Cover [4] point out, when Creighton’s Ants of North America [21] was published, dacetine ants were thought to be rare. It now appears that some dacetine ants, and in particular Pyramica species, are not uncommon. Rather, because of their cryptic appearances, diminutive size, and slow movements, colonies are frequently overlooked in the field. Although colonies are seldom collected, individuals in leaf litter samples have been reported throughout the world [22].
Since this is the first report of Pyramica colonizing artificial nests, it is premature to draw conclusions. At this juncture, we do not know how many different Nearctic species will colonize artificial nests. Artificial nesting cavities have been employed in several experiments performed in woodland settings. Herbers and Banschbach [24] successfully used hollowed-out wooden doweling to investigate nest choice by Myrmica punctiventris and Temnothorax longispinosus. The occupancy rate was as high as 27% per species. In another study on social organization of M. punctiventris, Herbers and Banschbach [25] used pieces of wooden doweling that had lengthwise holes. Friedrich and Philpott [26] used nests made from hollow twigs of differing internal diameters to study the impact of urbanization on cavity nesting ant communities in Ohio. Hollowed-out natural twigs were also used by Armbrecht et al. [27] to study the correlation between diversity of twig species versus the diversity of twig nesting ant species. They found a positive correlation. While other investigators have used artificial nesting cavities; we believe our nest design may be unique and favors some cavity nesting ant species over others.
4.2. Nests in Acorns
The decomposition of acorns follows a predictable series of steps from when the acorn first falls to the ground until it is incorporated into the humus of the forest floor [28]. This decomposition is aided by various insects and microorganisms, resulting in an empty outer shell. Depending upon the region, multiple species of weevil larvae (Curculionidae) and moth larvae (Tortricidae) feed on the seed inside the shell. The result is a partially empty hard shell with a small exit hole. Collection data document these degraded acorns are superb nest sites for a variety of ants including Pyramica. It is not clear how ants chose acorns for nesting. The volume of the cavity may be one parameter; others may include how the acorn decomposes and retains moisture, whether bacteria, fungi, or secondary plant substances are present, structural differences or biochemical differences.
Nest site selection by P. rostrata was originally described by L. G. Wesson and R. G. Wesson [8]. Several colonies were found in decaying logs. It is assumed that the nests were in preformed cavities. The authors report one colony in a rotten hickory nut. Talbot [18] also reported a P. rostrata colony from a hickory nut (Carya sp.) (Table 1). The Wessons [8] reported a colony found in a decayed stick in the leaf letter. Our data on P. rostrata suggest that the artificial nests may mimic natural cavities in rotten wood. Perhaps it is not the wood the ants are choosing but rather the environmental condition, that is, moisture content of the artificial nests. In retrospect, the reports of P. rostrata in hickory nuts could have given investigators a clue where else to look for P. rostrata colonies, namely, acorns. The data on colony demographics for P. rostrata recovered from acorns clearly documents the importance of acorns as natural nesting sites for this species.
Colonies of P. rostrata were also reported by L. G. Wesson and R. G. Wesson [8] living in the humus under the leaf layer. Another colony was found in a chamber in dry soil under a rock. Talbot [18] reports finding one colony in a chamber six inches below the surface (Table 1). It would appear that P. rostrata is either opportunistic or a generalist in choosing nest sites. As documented, this species constructs colonies in decaying logs, cavities in nuts and acorns, in the soil under a covering object, and in subterranean cavities.
Many myrmecologists have collected ants from acorns over the last half century, without finding Pyramica colonies. However, there is an anecdotal report by Coovert [29, page 95] that a colony of P. ohioensis was found in an acorn and state “Colony Organization: Futher data lacking.” In our study, fourteen colonies of P. ohioensis were recovered (Table 2). Two colonies were recovered from artificial nests, eleven from acorns and one from a hickory nut. At this juncture, it may be assumed that its nest site requirements are similar to those of P. rostrata. The differences in the number of nests recovered for P. rostrata and P. ohioensis may reflect the density of each or may indicate P. ohioensis has slightly different nesting requirements and that our two collecting methods do not adequately sample for this species.
4.3. Colony Structure
Colonies of some species of ants occupy more than one physical nest (polydomy) rather than one (mondomy). For example, Temnothorax longispinosus under laboratory conditions exhibits colony fission and fusion [30, 31]. Polydomy or forms of polydomy have been documented in T. curvispinosus [32], T. ambiguus [33], Myrmica punctiventris [34, 35], and Stenamma diecki [35, 36].
Worker number in a colony changes with the annual cycle. Our data represent the period after the reproductive alates have swarmed and just before winter hibernation. These are the first data sets based on multiple collections for each species. Pyramica rostrata colonies exhibited a mean of eighty-three workers per colony, compared to fifty-six workers per P. ohioensis. Our data are similar to those worker numbers per colony reported for other Nearctic species of Pyramica as listed in Table 3. The Pyramica pergandei colony reported by Brown [37] from Massachusetts that contained seven hundred workers is markedly larger than those listed in this study. Obviously we have much to discover about worker numbers in Pyramica colonies.
It is difficult to compare our Pyramica data to other data sets since so little exists for the Nearctic region, or that matter worldwide. Dejean (reported by Beckers et al. [44]) provides data on colony worker number for four African species: Pyramica emarginata [199 workers], P. lujae [57 workers], P. serrula [78 workers], and P. truncatidens [133 workers]. The number of workers per colony for the Japanese ground nesting “tramp species” Pyramica membranifera, is 198 ± SD 129, [45]. Our data seems to fit in the range reported by these authors.
Larvae were present in most colonies of P. rostrata and P. ohioensis. In acorns P. rostrata colonies had a mean of 63.3 ± SD 38.0, larvae and P. ohioensis has 39.4 ± SD18.9, . Figures 4 and 5 plot workers as a function of the number of larvae present for each colony recovered from acorns by species.
(a)
(b)
(a)
(b)
4.4. Polygyny
The mean number of queens per colony for P. rostrata is 5.0 ± SD 4.9, , whereas P. ohioensis had a mean of 2.6 ± SD 1.7, queens per colony. Thus it appears P. ohioensis forms smaller colonies and has fewer queens per colony.
Analysis of the twenty-eight P. rostrata colonies and fourteen P. ohioensis collected in the late summer documents that most colonies contain multiple queens; only three colonies did not. It is not clear whether these are true monogynous colonies where each colony was founded by a single queen and is retained as a single queen throughout the life of the colony. These colonies could have been established by a single queen and a group of workers leaving a larger polygynous colony.
Polygyny is an important component of ant biology and has evolved independently many times [46]. Pyramica rostrata and P. ohioensis are part of a growing list of polygynous species. It is not clear how many queens were inseminated, if one queen is doing the egg laying or if it is a shared responsibility. Given that the number of queens in both P. rostrata and P. ohioensis colonies does not correlate with the number of larvae or number of workers, it appears that only one queen is laying eggs. Polygynous colonies can be formed by swarming where a single queen or multiple queens leave the parent colony with a small group of workers and establish a new colony. Colonies could also be established by a single queen with extra queens added as the colony matures.
Five colonies of P. rostrata and three of P. ohioensis contained at least one male and/or female alate. The end of August seems relatively late in the year for alates to be present in the colonies. Although one assumes that the female individuals would eventually leave the colony, they may not. Ito et al. [45] working with the polygynous species Pyramica membranifera recently reported thelytokous reproduction by queens. In laboratory investigations, they documented that some alate queens found in the colonies, shed their wings and established new polygynous colonies which produced workers.
4.5. Utility of Artificial Nests
This investigation demonstrates the limited effectiveness of a novel artificial nest to obtain colonies of two species of Pyramica. Additional work is necessary to determine whether other species of Pyramica will inhabit artificial nests, perhaps based on this design, if the nests are placed in a suitable habitat. The capture and subsequent maintenance of colonies of Pyramia in the laboratory will allow a number of important questions about the biology of Pyramica to be addressed. As pointed out by Deyrup and Cover [4], the natural history of these ants is poorly understood. We hope that our contribution will allow other investigators an opportunity to employ our methods to pursue investigations of this most fascinating group of ants.
Acknowledgments
The authors would like to thank Donna Maglott for assistance in multiple aspects of this work. Discussions and a paper review by Stefan Cover improved the quality of this publication. They appreciate the comments and information provided by the reviewers which helped strengthen and improve the paper. This publication acknowledges the pioneering work on dacetines in North America by William L. Brown, Jr. and Edward O. Wilson.