Foundress behaviors and interactions in polygynous colonies of the haplometrotic Japanese paper wasp Polistes snelleni (Hymenoptera: Vespidae)

INTRODUCTION

Colonies of Polistes wasps are initiated by a single foundress or multiple foundresses. Most species studied in the genus Polistes show both types of colony founding (Reeve 1991), and the proportion of each type varies from species to species or even from population to population (Itô 1993). However, the reason why these differences occur is not well understood.

In the Japanese Polistes species, the frequency of multiple-female colony founding (pleometrosis) is quite low compared with that in North American and European species, even at similar latitudes (Itô 1993). Why single-female colony founding (haplometrosis) dominates in the Japanese species is a very interesting question, as is the converse question, that is, why does multiple-female colony founding occur only rarely?

There have been several reports on polygynous colonies in Japanese species of the subgenera Polistes (Yamane 1973; Hoshikawa 1979; Makino & Aoki 1982; Kojima & Hagiwara 1990) and Gyrostoma (Yoshikawa 1957; Kasuya 1981; Takamizawa 1982; Hirose & Yamasaki 1984; Tsuchida & Itô 1991). In the subgenus Polistella, only two cases of polygynous colony are known: one in Polistes snelleni de Saussure (Yamane 1969) and the other in Polistes nipponensis Pérez (Y. Sakamoto, unpubl. data, 1993). These reports showed that unlike in the North American and European species, clear division of labor is not found, even if there seems to be dominance behavior among foundresses in these colonies.

Given that multiple-female colony founding is very rare in the Japanese Polistes, analyses of foundress behaviors and interactions in multiple-foundress colonies may help us understand the evolution of single-female colony founding. Here, I report the frequency of polygynous colonies, and foundress behaviors and interactions in polygynous colonies of the Japanese paper wasp P. (Polistella) snelleni.

MATERIALS AND METHODS

The study was conducted at Misumai, Sapporo, in northern Japan (42°58′N, 141°16′E) between 1989 and 1996. Every year, as many P. snelleni nests as possible in the egg to early larval substages (the most advanced brood being eggs or young larvae) initiated by overwintered females (foundresses) were found in an area of approximately 8 ha in spring. Nests were usually checked at least every 2 days to determine whether they contained one or more foundresses. In the calculation of the frequency of polygynous colonies, colonies within 1 month after the start of the nesting season were counted.

Foundress behaviors were observed for a total of 29.7 h during the 9 days between 29 May and 18 June in a polygynous colony found in 1993. Behavioral observations were made between 10.00 and 15.00 hours. This period corresponded to the second half of the egg substage and the early larval substage in the development of this colony. Each foundress was marked with quick-drying colored paint for identification. For each foundress, the time and number of take-offs from and landings on the nest were recorded. When foundresses landed on the nest, they were checked to record what kind of load they had brought. Loads observed were prey for food and pulp for nest material. Oviposition, egg-eating (oophagy), and abdomen-rubbing on the nest pedicel or comb surface were also recorded. When both foundresses were on the nest, interactions such as deliveries of collected material, deliveries of regurgitated liquid (trophallaxis), approaches to the other foundress, and touching the other foundress with antennae (antennation) were also noted.

RESULTS

Foundress behavior in one polygynous colony in 1993

In the 1993 polygynous colony, both foundresses foraged and laid eggs. Although there were no significant differences between the foundresses in terms of the performance of several tasks, with the exception of prey collection (two-tailed Mann–Whitney U-test, U = 13.5, P < 0.02, n₁ = n₂ = 9), one foundress (foundress A) left the nest and worked more frequently than the other (foundress B) (Table 1). Although the foundresses’ body sizes were not measured, there was no apparent difference.

Table 1. Mean number of various behaviors per hour (± SD) on 9 observation days

Behavioral category	Mean no. times/h (± SD) (n = 9)		Mann–Whitney U-test
	Foundress A	Foundress B
Foraging trip	1.43 ± 0.98	0.76 ± 0.62	NS
Prey collection	0.21 ± 0.17	0 ± 0	P < 0.02
Pulp collection	0.47 ± 0.54	0.28 ± 0.43	NS
Oviposition	0.16 ± 0.28	0.10 ± 0.15	NS
Oophagy	0.04 ± 0.11	0 ± 0	NS
Rubbing on  pedicel	1.67 ± 1.04	1.81 ± 0.89	NS
Rubbing on  comb	0.23 ± 0.43	0.20 ± 0.31	NS

• NS, Not significant.

With respect to foraging activity, only foundress A collected insect prey (Table 1). Both foundresses collected pulp, and they made new cells or elongated the existing cell walls. Egg-laying by both foundresses was observed. Foundress A initiated new cells in which the eggs were laid by foundress A and foundress B. Although foundress B initiated one new cell, I did not observe foundress B laying an egg into it. On one occasion, foundress A ate one of foundress B’s eggs just after it had been laid. Although the egg of foundress B that was eaten by foundress A was replaced with a new egg by the next day, I did not observe which female laid the egg.

Abdomen-rubbing behavior is interpreted as a way of applying ant repellent substance (Jeanne 1970; Kojima 1993). No differences were found in rubbing behavior between foundresses; both rubbed not only on the nest pedicel but also on the comb surface. Rubbing on the pedicel was more frequent than on the comb surface (Table 1).

Although the nest was usually attended by one or both foundresses during the observation period, absence of both wasps from the nest sometimes occurred (Fig. 1).

DISCUSSION

My results, except for the absence of aggressive dominance behavior between foundresses, are almost the same as those reported for other Japanese Polistes species. The frequency of polygynous colonies was low, and a dominance hierarchy was not established between the foundresses. Although Yamane (1969) reported the presence of one polygynous colony of P. snelleni among more than 100 colonies observed, he could not find any cooperative or dominance behavior between the two foundresses. Behavioral interactions, including domination among foundresses, are important to establish a dominance order and “subordinate workers” (Hunt 1991) (Pardi 1948; West-Eberhard 1969). Makino and Aoki (1982) divided Japanese polygyny into two types on the basis of the presence or absence of domination behavior. According to their classification, domination did not occur in the 1993 polygynous colony in the present study. Regardless of domination, however, oviposition or the development of ovaries occurs in every foundress of the Japanese polygynous colonies (Kasuya 1981; Makino & Aoki 1982). Both foundresses on the studied colony laid eggs during the observation period (Table 1).

Not only in the larval substage but also in the egg substage when no larvae were present in the nest, foundress A collected prey and brought it to the nest. Moreover, after landing on the nest, foundress A attempted to give the prey to foundress B (Table 2). However, foundress A showed reproductive domination by eating one of foundress B’s eggs (Table 1). These behaviors of foundress A are inconsistent with those of dominant females in North American and European species, in which dominant foundresses monopolize oviposition and rarely forage (Pardi 1948; West-Eberhard 1969).

Colony survival (Gibo 1978; Reeve 1991; Itô 1993) and nesting habits (Klahn 1979; Cervo & Turillazzi 1985) have been suggested to be responsible for polygyny or pleometrosis. In contrast, dispersion of foundresses from their natal nest sites (West 1968; Yamane 1986; Suzuki & Ramesh 1992) is thought to promote monogyny or haplometrosis. However, these factors have not been critically examined to show differences in the formation of foundress associations among populations or species.

Why polygynous colonies were formed in only two colonies of P. snelleni is unclear. In this population, colony usurpation frequently occurs in the second half of the pre-emergence stage (Makino & Sayama 1991). Although multiple-female colony founding should be more advantageous in such a situation (Gamboa 1978), its frequency is still low. The abundance of substrates such as twigs and a weak tendency toward philopatry (Makino et al. 1987) may contribute to the maintenance of single-female colony founding. In any case, to understand the maintenance and evolution of single-female colony founding, it is still necessary to gather data on foundress behaviors and interactions in polygynous colonies of the typically haplometrotic Polistes species.