Taxonomy, distribution and nesting biology of species of the genus Dolichovespula (Hymenoptera, Vespidae)

Sources of species data

Species habitat and nest site characteristics

Dolichovespula adulterina is a social parasite of D. arenaria and D. alpicola in the Nearctic region and D. saxonica in the Palearctic region.

Dolichovespula alpicola is associated with mountainous regions, although it is found at lower altitudes in Alaska, being associated with mixed hardwood/coniferous forests. One colony has been found in a wall void and a second colony on the surface of the ground surrounded by leaf litter. Label data from specimens indicate that it is mainly found at altitudes between 1700 m and 3300 m a.s.l.

Dolichovespula arenaria is found in a wide range of habitats from sea-level to approximately 3350 m a.s.l. Nest sites are usually aerial in trees and shrubs up to 17 m above the ground, but also in bird boxes, under rock underhangs and in decayed logs. Nests may also be found in leaf litter or at a shallow depth in the soil. Workers have been observed to remove soil particles.

No information is available on the nest site characteristics of D. baileyi. Label data from specimens indicate that it has been found at altitudes between 1300 m and 2500 m a.s.l.

No information is available for D. flora.

Dolichovespula intermedia is associated with open habitats in mountainous areas including subalpine pasture, maize terraces, subalpine scrub, alpine steppe and forest. It has been recorded at altitudes between 1200 m and 4120 m a.s.l. A young colony with three workers was found in an aerial situation under the eaves of a hut (Archer 1987).

No information is available for D. lama, but label data indicate that it occurs in mountainous regions at altitudes between 3350 m and 4200 m a.s.l.

Dolichovespula maculata is associated with forested areas or vegetation in urban areas. Nests are aerial usually in trees and shrubs but also under rock underhangs and on buildings, (e.g. under eaves, gutters and porches and against glass windows). Nests have been recorded 1.1 m to 20 m above ground level.

Dolichovespula media is found in a wide range of habitats, with aerial sites for its nests, for example usually trees, shrubs and under the eaves of buildings from approximately 1 m to 5 m above ground level.

Dolichovespula norvegicoides is associated with closed coniferous forest. Nests are usually aerial in shrubs and trees up to 6.7 m above ground level. Some nests found in shrubs can be at a very low level, so that they are partly in the leaf litter. Nests have also been found in decayed logs, in a cavity of a stream bank, within a root tangle of fallen trees, in wall voids and on the underside of a plywood sheet.

Dolichovespula norwegica is found in a wide range of habitats. Nest sites are usually aerial in trees and shrubs up to 10 m above ground level, but can be found partly in the leaf litter layer. They are also found in a range of miscellaneous sites, such as a hole in the ground, the eaves of a house, on a wall, on a large stone, in a haystack, attached to a window frame, in a beehive and under a garden seat.

Dolichovespula omissa is a social parasite of D. sylvestris.

Very little is known about the nest site characteristics of D. pacifica, but it has been found in hollow trees and concealed places above and below ground.

No information is available for D. panda, but label data indicate that it occurs in mountainous areas at altitudes between 1000 m and 4300 m a.s.l.

Dolichovespula saxonica usually nests in aerial situations in shrubs and trees up to 2 m above the ground. The species is also found nesting in beehives, under eaves of buildings up to 7 m above the ground, under porches, in wall cavities and rarely in a tree hole or in a hole in the ground.

No information is available for D. stigma, but label data indicate that it occurs in mountainous regions at altitudes between 1000 m and 3000 m a.s.l.

Dolichovespula sylvestris nests under overhanging banks, in shallow holes in the ground when the nest is usually visible, at the leaf litter/mineral soil interface and more rarely in shrubs and trees. The workers are able to remove earth particles to enlarge a nesting cavity. Nests have also been found in cavities of old walls, under eaves and porches, in outhouses, in straw bales, in roof spaces and beehives.

No information is available for D. xanthicincta, but label data indicate that it occurs in mountainous regions at altitudes between 4000 m and 5470 m a.s.l.

Types of colonial data

The development of a colony can be described using three types of data: (i) colonial characteristics (i.e. the number of cells, brood stages, adults and meconia); (ii) developmental parameters (i.e. rates of cell building and egg-laying, and the adult production of each caste and sex); length-of-life characteristics (i.e. the length of each brood stage, of workers and of the colony itself).

Ideally these three types of data should be obtained from the inception to the death of a colony, and studied under natural conditions. In practice, such data are difficult to obtain because colony contents are covered by envelopes, the nests are often in enclosed sites, and the wasps are aggressive when colonies are investigated.

Another approach is to collect colonies throughout the year, make counts of the colonial characteristics, and use laboratory colonies to determine the length-of-life characteristics. From these characteristics, estimates can be made of the developmental parameters. This second approach has been used as the main source of data in the present paper.

Localities where colonies were sampled

Development phases of colonies

Colonies can be arranged in groups according to the phase of their development. Three phases are recognized:

The large cell phase is divided into four subphases, with the last one further divided into three subphases when there are sufficient data:

Recognition of subphase LCCAB requires meconia counts.

Critical dates in the development of colonies

The critical dates in the development of a successful colony are nest foundation by a queen, emergence of the first workers, start of large-cell building and adult queen emergence and termination of the colony.

Critical dates can be determined by direct observation of events or indirectly either by forward or backward calculation from colonies at a known developmental phase or from a brood with a stable age distribution. A stable age distribution of the brood is reached when the proportions of the brood stages (eggs, larvae, sealed brood) show little change with time. Then the percentage of each brood stage would be proportional to their duration.

For forward and backward calculation it is necessary to know the length of time of the egg, larval and sealed brood stages. For example, from a LCCL colony the start of LCC and LCCA can be estimated as follows. The start of LCC would be, on average, the duration of the egg stage plus half of the duration of the larval stage before the date of collection, whereas the start of LCCA would be, on average, half of the duration of the larval stage plus the duration of the sealed brood after the date of collection. Forward and backward procedures cannot be used for LCCA colonies.

For the length of the queen colony there are a few direct observations (Table 2A) that indicate a mean length of 23–30 days, although most data indicate 23–24 days. The direct observational data of Brian and Brian (1952) is based on daily cell counts of five colonies that succeeded in producing workers. From direct observation of nest foundation and emergence of the first worker of D. sylvestris (n = 14), Archer (2000) estimated that the queen colony, on average, lasted 24 days. There are also few direct observation data for the lengths of the egg, larval and sealed brood stages (Table 2A). Allowing a queen colony development time of 24 days, on average, the egg stage would last 5 days, the larval stage 9 days and the sealed stage 10 days.

Again, there are few direct observations (DO) of the length of brood stage for the small-cell and large-cell colonies, so that use of estimations based on “stable-age distributions” (SAD) and “forward and backward calculations” (FBC) are needed (Table 2B). The use of 5 days for the length of the egg stage for SAD and FBC is derived from five directly observed field colonies of D. sylvestris (Brian & Brian 1952). Combining the laboratory data, the mean lengths of brood stages are: egg 3.8 days, larva 10.9 days, sealed brood 10.8 days, giving a total duration of 25.5 days. Combing the field data, the mean lengths of the brood stages are: egg 5 days, larva 10.6 days, sealed brood 11.3 days, giving a total duration of 26.9 days. The mean lengths of the brood stages reared in small cells varied between 27 and 32 days (mean 29.6 days, egg stage 5 days, larval stage 11.3 days, sealed brood stage 13.3 days) and those reared in large cells varied from 21 to 36 days (mean 26 days, egg stage 5 days, larval stage 10 days, sealed brood stage 11 days).

Colony cycles

In the USA, in Washington State, nest initiation of D. maculata is usually from mid-May, with the first workers emerging in mid-June, large-cell building starting from mid-July, the first queens emerging during mid-August, and colony termination occurring during mid-September The life cycle is approximately 4 months (122 days). At the lower latitude of Indiana, colonies can be initiated earlier by early May and terminated later during late September, giving a life cycle of approximately 5 months (153 days). At the still lower latitudes of central California, initiation can be from late March, but at the higher latitude of British Columbia, initiation is delayed until early June. Colonies in central California may have a 5–5.5-month (155–170 days) life cycle, with colonies terminating in late August or early September.

In England, colonies of D. media are initiated from early May, with the first workers appearing from early June, large-cell building starting from early July, the first queens emerging from early August, and colonies terminating by the end of August. The colony cycle is approximately 3.3 months (112 days). At the higher latitude of Hokkaido, Japan, colony initiation is delayed until the end of May, with colony termination during early September, giving a colony cycle of approximately 3.2 months (97 days).

In England, colonies of D. norwegica are initiated from mid-May, with the first workers appearing from early June, large-cell building starting from mid- to late June, the first queens emerging from early July, and with colonies terminating from mid-August, although with favorable weather, some colonies persist until early September. The colony cycle is 3.1–3.7 months (95–115 days).

Few data from England are available for D. saxonica, but probable dates are as follows: colony initiation just before mid-May, the first workers emerging during early June, large-cell building starting by mid-June, first queens emerging from early July, and colony termination soon after mid-August, although with favorable weather, colonies may continue longer. The colony cycle could vary between approximately 3 and 3.7 months (91–111 days). In Hokkaido, Japan, colony initiation is delayed to mid-May or later, with the first workers emerging during mid-June, large-cell building starting from late July, the first queens emerging from late August, and with colony termination by early September. The colony cycle is approximately 3.5 months (109 days).

In north-west USA, particularly Washington State, colony initiation for D. arenaria starts from late May, with the first workers appearing from mid-June, large-cell building starting from early July, the first queens emerging from late July, and colony termination by early September. The colony cycle is 3.5 months (105 days). At the lower latitude of the southern part of central California, colony initiation starts earlier, from late March, and, according to Greene et al. (1976), terminates at the end of July, with a colony cycle of 4.3 months (132 days). Duncan (1939) indicated that colony termination occurs during late August, giving a colony cycle of approximately 5.3 months (163 days).

In England, colonies of D. sylvestris are initiated from mid-May, with the first workers emerging from early June, large-cell building starting from late June, the first queens emerging from mid-July, and colony termination by the end of August, or, with favorable weather, during early September. The colony cycle is 3.5–3.8 months (107–117 days).

The queen of the social parasite D. adulterina invades a host colony when the first workers are beginning to emerge. The inquiline queen coexists with the host queen for approximately 10 days before the host queen is killed. The inquiline queen lays her eggs from early July or earlier, with males emerging from late July and queens from early August. The males and queens remain in the host colony for a few days before leaving on their mating flights. The remaining workers then usually succeed in rearing some host males before the colony terminates from mid-August. The inquiline queens usually die by mid-July, but may die earlier as a result of injuries received from fights with the host workers. If the inquiline queen dies earlier, she may only succeed in laying some male eggs. The colony cycle for the successful inquiline queen is approximately 2.8 months (86 days). Much less is known about the social parasite D. omissa, but its males can emerge from late June and its queens from late July.

Little information is available for the remaining 10 species, often only label data relating to the sex or caste of the specimens. For the mountainous species, colony initiation seems to be delayed and adult activity extended. For D. intermedia and D. alpicola, colony initiation is delayed until late June or July. Workers are active until October for D. alpicola, until November for D. pacifica and D. xanthicincta, and even until December for D. stigma. Male activity also extends to October for D. alpicola and November for D. pacifica. New queen activity is also extended until October and November for D. xanthicincta.

Phase 0: The queen colony

The available data on the size of mature queen colonies are given in Table 3. Mature queen colonies considering data from all species have a mean of 40 cells (range 24–61 cells for different colonies). The variation in cell building by a queen reflects her activity and may be used as a measure of queen quality. The mean number for each brood stage of the mature queen colony is 12 eggs, 19 larvae, eight sealed brood and two empty cells (i.e. 41 cells). The small difference between the estimates of the mature queen nest based on cell and brood counts is due to sample size difference. The number of cells for each species is very similar. There is an accumulation of larval cells with a smaller similar number of egg and sealed brood cells. The number of sealed brood cells can be used as a measure of queen quality, because it represents the ability of the queen to bring larvae to maturity so that pupation becomes possible. The accumulation of larvae is an indication of the queen’s inability to bring larvae to maturity.

Phases 1 and 2: The small cell and large cell colonies

Colony samples were often small or not available for some phases and subphases of each species. A decision was taken that generally only phases and subphases for each species with three or more samples would be considered. As such, the data for each phase or subphase were derived from the following species: SCC: D. maculata, D. media, D. norwegica, D. saxonica, D. sylvestris; LCCE: D. maculata; LCCL: D. maculata, D. media†, D. norwegica†, D. sylvestris†; LCCS: D. maculata, D. media, D. norwegica, D. saxonica, D. sylvestris; LCCAA: D. norwegica, D. sylvestris, D. norvegicoides; LCCAB: D. maculata‡, D. media‡, D. norwegica, D. sylvestris; LCCAC: D. adulterina, D. maculata, D. media, D. norwegica, D. saxonica, D. arenaria, D. sylvestris; where † is LCCE and LCCL data combined, and ‡ is LCCAA and LCCAB data combined.

Number of cells

Table 4 shows the mean mature colony size as the number of cells for each species of Dolichovespula. Data are available for eight species, although data for D. norvegicoides are derived from LCCAA colonies, with two sets of data for D. media, D. saxonica and D. arenaria. The two data sets for D. media and D. saxonica are very similar to each other. There are fewer cells for D. arenaria from Oregon than for those from Washington State and those for all other species. Overall, the mean mature colony size for the seven non-parasitized species is 1087 cells. The colonies of D. arenaria parasitized by D. adulterina are much smaller, with a mean of 299 cells.

Table 4 also shows the sizes of the smallest and largest colonies that have been found. Larger colonies have been found in other data sets: D. maculata, 3300 cells from Indiana (MacDonald & Deyrup 1989); D. media, 1724 from Nagano, Japan (Takamizawa 1981); D. saxonica, 2148 cells from Germany (Kemper & Döhring 1967); D. arenaria, 2489 cells from California (Duncan 1939); D. sylvestris, 2108 cells from England (Archer 2002); D. arenaria-adulterina, approximately 650 cells from Massachusetts (Evans 1975). In conclusion, nests usually reach approximately 1000 cells, but on rare occasions can reach 2000–3000 cells.

Table 4 also shows the mean number of small and large cells in mature colonies. Determination of the number of small and large cells is complicated by the presence of mixed combs of small and large cells in some colonies. In mature colonies of D. maculata, the first or oldest comb, and the second and third combs may be mixed combs. Akre and Myhre (1992) did not usually make separate counts of the small and large cells on these mixed combs.

In Japan, Makino (1982) did not find mixed combs in D. media colonies. In contrast, Archer (the present paper) found that three colonies out of 21 colonies (14.3%) contained small and large cells on the second comb. These small cells represented 81.8, 33.1 and 81.8% of the cells of the second comb.

In colonies of D. arenaria, Roush and Akre (1978) found that the first and second and sometimes the third combs were mixed combs. Greene et al. (1976) also found colonies with mixed first, second, third and one colony with a fourth comb, although usually only the first and second combs were mixed combs. On the mixed combs the large cells were peripheral to the central small cells. Roush and Akre (1978) and Greene et al. (1976) did not make separate counts of the small and large cells on mixed combs.

In colonies of D. norwegica, Archer (2000) found a mixed second comb in four colonies from a sample of 28 colonies. The percentages of small cells on the second comb were estimated as 7, 3, 5 and 4%, giving a mean of approximately 5%. In colonies of D. saxonicaMakino (1982) and Archer (the present paper) found no colonies with mixed combs. Akre and Bleicher (1985) and Akre and Myhre (1994) found mixed combs in three out of nine colonies of D. norvegicoides. Finally, Archer (2000) found colonies of D. sylvestris with a mixed first comb in 18 out of 28 colonies. These first combs consisted of approximately 5% large cells. One colony was found with mixed first and second combs.

The number of cells increases slowly during SCC, with a mean for all species of 139 cells (range of means for different species: 83–224 cells), before increasing to a mean for all species of 321 cells (range of means for different species: 231–568) on 2.2 combs during LCCE and LCCL, and to a mean for all species of 513 cells (range of means for different species: 444–677) on 2.6 combs during LCCS. With the emergence of large-cell adults cell building continues, but eventually ceases at a mean for all species of 1087 cells (range of means for different species: 778–1643) on 3.6 combs.

Data relating to the small and large cells are limited to colonies of D. media, D. norwegica, D. saxonica, D. norvegicoides and D. sylvestris. From SCC the number of small cells rapidly increases to a mean for all species of 290 cells (range of means for different species: 251–313) on one comb by LCCS and then to a mean for all species of 338 cells (range of means for different species: 294–431) on 1.1 combs during LCCA. During LCCL the mean number of large cells for all species is 44 cells (range of means for different species: 24–50) on one comb, increasing to a mean for all species of 281 cells (range of means for different species: 176–364) on 1.6 combs during LCCS before reaching a mean for all species of 831 cells (range of means for different species: 484–1303) on 2.7 combs.

The small cell/large cell ratios vary from 0.4 to 0.9 between species (Table 4). Such values indicate that more large cells are built than small cells. Plots for mature colonies of each species of the number of small cells divided by the number of large cells versus the total number of cells show negative relationships. For three species for which there are sufficient data, these negative relationships are statistically significant (Table 5), indicating that as colonies become larger, a proportionally greater number of large cells are built. Similar negative relationships also were found for D. media and D. saxonica by Makino (1982).

Cell building rates per day can be estimated by dividing the number of recently built cells (i.e. empty and egg cells without meconia) found in a colony by the development of the egg stage, assumed to be 5 days. The cell building rate per day per worker can be estimated by dividing the cell building rate per day by the number of workers. Such estimates are only possible from colonies where meconia and egg counts have been made (i.e. colonies of D. norwegica and D. sylvestris). The highest mean small-cell building rates per day are during LCCL (7.6) and LCCAA (11.7) for D. norwegica and during LCCL (9.4) for D. sylvestris. The highest mean large-cell building rates per day are during LCCAA (46.1) and LCCAB (49.4) for D. norwegica and during LCCAA (21.8) for D. sylvestris. There are insufficient data to make an estimate during LCCAB for D. sylvestris. During LCCAC, cell building rates are at or near zero. The highest mean cell building rates per day per worker are during SCC and LCCL, when the rate varies between 0.5 and 1.1. Later it decreases, varying between 0.26 and 0.27 during LCCS, between 0.12 and 0.36 during LCCAA and down to 0.17 during LCCAB.

Brood

The number of brood cells, among species, increases with the increase in the number of cells up to LCCAB, when there is a rapid decline. Egg-laying rates per day can be estimated by dividing the number of eggs found in a colony by the developmental time of the egg stage, assumed to be 5 days. These estimates were derived from colonies of D. media, D. norwegica, D. norvegicoides and D. sylvestris. The egg-laying rate per day in small cells increases from a mean for all species of six eggs during the SCC to a mean for all species of 18 eggs during LCCS until LCCAB before decreasing to, or near to, zero during LCCAC. Egg-laying rates per day in large cells increase from a mean for all species of six eggs during LCCL to a mean for all species of 18 eggs during LCCS and to a mean for all species of 36 eggs during LCCAA and LCCAB before decreasing to, or near to, zero during LCCAC.

Adults

During SCC the number of workers is small, with a mean for all species of 14 (range of means for different species: 5–45). The number of workers then increases to a mean for all species of 52 (range of means for different species: 37–145) during LCCL and a mean for all species of 79 (range of means for different species 24–104) during LCCS, reaching a mean for all species of 170 (range of means for different species 67–294) during LCCAA and LCCAB, before rapidly decreasing during LCCAC.

Individual colonies have been recorded with many more workers: D. maculata with 771 from Indiana, D. norwegica with 363 during LCCAB from England, D. arenaria with 697 from Washington State and 695 from British Columbia, and D. norvegicoides with 326 during LCCAA from Washington State.

By making meconia counts within small cells, estimates can be made of worker survival during SCC and the early subphases of LCC. The best available data are for D. sylvestris. During SCC worker percentage survival varies between 32 and 100%, during LCCL between 55 and 90% and during LCCS between 46 and 95%.

During the early subphases of LCC, plots of the number of cells versus the number of workers show a positive relationship, but because of a small sample size, no statistically significant relationships can be found except for D. sylvestris. For D. sylvestris, during LCCL the correlation coefficient is 0.86 (n = 11, P < 0.001). Plots of the number of small cells divided by the number of large cells versus the number of workers have a negative relationship, but again, except for D. sylvestris, sample sizes are too small to produce statistically significant relationships. For D. sylvestris, during LCCL the correlation coefficient is 0.78 (n = 11, P < 0.01). These relationships indicate that the more workers that are present in a colony during LCCL, the more cells, and proportionally more large cells, are built. Worker number and survival in the early colony is important for the future development of the colony.

There have been no direct observations of the durations of the lives of workers. An estimate can be made when the brood stages show a stable age distribution, so that the numbers of brood and adults would be proportional to their durations. Using the small-cell brood, estimates are possible for three species: D. media (19 days), D. norwegica (12 days), and D. sylvestris (15 days), giving an overall mean of approximately 15 days.

The larva/worker ratio is an indicator of the amount of food each larva receives. The mean ratio for all species is 5.9 (range of means for different species: 5.5–7.0) during SCC, then decreasing to a mean for all species of 4.4 (range of means for different species: 4.2–4.4) during LCCL and then to a mean for all species of 3.1 (range of means for different species: 3.0–3.2) during LCCS. During LCCAA and LCCAB, the ratio reaches its lowest value with a mean for all species of 2.3 (range of means for different species: 1.8–2.7). The ratio is at its lowest when the majority of the large-cell brood is being reared.

The rates of adult emergence per day can be estimated by dividing the number of sealed brood by the duration of life of the sealed stage (Table 2B). The mean adult emergence rates per day from the small cells for all species are 1.9 adults (range of means for different species: 0.7–2.5) for SCC, increasing to a mean for all species of 6.0 workers (range of means for different species: 4.1–7.8) for LCCL, to a mean for all species of 12.0 adults (range of means for different species: 10.7–12.6) for LCCS, to a mean for all species of 13.7 adults (range of means for different species: 13.4–14.1) for LCCAA and decreasing to a mean for all species of 6.8 adults (range of means for different species: 4.6–10.0) for LCCAB. Mean adult emergence rates per day from the large cells for all species are 4.7 adults (range of means for different species: 3.9–6.4) for LCCS, increasing to a mean for all species of 19.7 adults (range of means for different species 19.3–20.1) for LCCAA and a mean for all species of 25.6 adults (range of mean between species 11.1–47.3) for LCCAB. During LCCAC adult emergences are zero or very small, as many sealed brood die in their cells.

Males are produced in both the small and large cells as shown by the presence of male and worker sealed brood in these cells. For D. sylvestris, the percentage of small cells used to rear males is 0% until LCCS, when it is 15.9%, increasing to 55.9% during LCCAA and 73.4% during LCCAB. The percentage of large cells used to rear males is very similar during LCCS (39.2%), LCCAA (34.9%) and LCCAB (44.6%).

For D. norwegica, the percentage of small cells used to rear males is 0% until LCCS (21.2%) and LCCAA (9.1%), before increasing during LCCAB (48.4%). Like D. sylvestris, the percentages of large cells used to rear males are similar during LCCS (54.2%), LCCAA (60.5%) and LCCAB (53.3%). The production of males from the large cells in colonies of D. norwegica differs from their production in D. sylvestris, in that each colony of D. norwegica rears either mainly male or queen brood.

The specialization of colonies in producing males or queens in their large cells has also been found in D. arenaria (Greene et al. 1976), although a colony may show both male and queen phases. In 15 colonies (47% of the sample), phases of both male and queen brood were present, usually with a queen phase preceding a male phase. In 10 colonies (31%), only the male phase was present, and in seven colonies (22%) only the queen phase was present. In the 25 male phases recorded, male brood represented 99% (83–100%) of the brood and in 22 queen phases, queen brood represented 97% (87%−100%) of the brood. Male production in small cells by queen phase colonies was considered to be very small; otherwise males were reared in 22.5% of small cells.

Adult production

The total number of adults produced from the small and large cells of a colony can be estimated by counting the meconia at the bottom of each cell. Small cells are used to rear one, two or three, but rarely four generations. At first workers are produced but later male production occurs. Large cells are usually only used once to rear one generation of either queens or males. If a second generation of males is started they usually die with the termination of the colony and only on rare occasions do these males become adults.

Table 6 shows for each species the number of adults produced from the small, large and all cells. The mean total adult production between species varies between 690 and 1581 adults. The largest adult production varies between 1090 and 3333 adults, although in excess of 6500 adults have been recorded from a colony of D. arenaria (Greene et al. 1976). Akre and Myhre (1992) found that the mean total adult production from six selected colonies of D. maculata was 937 adults, with 1911 adults as the largest production from a single colony.

Table 7 shows the mean cell use for each species for adult production. Overall each small cell produces 1.27–1.50 adults and each large cell 0.39–0.64 adults, or all cells are used between 0.67 and 0.88 times. From six selected colonies, Akre and Myhre (1992) estimated that total cell use varies between 0.79 and 1.19. For two particularly large colonies of D. arenaria, Greene et al. (1976) found total cell use of 1.57 and 1.92 adults. Akre and Myhre (1992) also noted from their selected six colonies that larger colonies produced relatively more adults per cell.

Plots of the number of adults produced from small-, large- or all cells versus the number of small-, large- and all cells show positive relationships that are statistically significant for three species (Table 8).