Multifemale nests and rudimentary castes in the normally solitary bee Ceratina japonica (Hymenoptera : Xylocopinae)

Abstract

Ceratina jap?nica is most commonly a solitary bee. About 20% of nests contain more than one female. Multiple female nests are usually reused from a previous season. Ovarian development of cohabiting females is variable and about one-half of them are uninseminated. Artificial induction of multifemale nests was attempted. In 11 out of 12 such nests, the larger females served as guards and the smaller females as foragers. Eggs were laid either by the larger females alone or by both larger and smaller cohabitants. Apparently C. jap?nica is at an incipient stage of social evolution. Since Hamilton's studies (1964), theoretical explanations of the origin and development of sociality in Hymenoptera have received much attention, resulting in the publication of many interesting articles (for reviews see Wilson [1971] and Starr [1979]). In order to improve our knowledge of these problems, we need, besides a theoretical framework, much more precise information on the lives of primitively social wasps and bees (West Eberhard, 1978), especially on those occurring at the boundary between solitary and social life. The present paper deals with the occurrence of rudimentary castes in multi female nests (MFN) of the basically solitary bee Ceratina (Ceratinidia) jap?nica Cockerell. The small carpenter bees, Ceratina, have long been considered solitary (Michener, 1974). Chandler (1974), however, published a brief note suggesting the presence of some social tendencies. In certain Japanese species traits exist that can be regarded as precursors of social life, and MFN are occasionally found in the brood rearing season (Sakagami and Maeta, 1977). This paper gives further information on (1) the analysis of MFN found in nature and (2) artificial induction of such nests. Materials and Methods life of c. jap?nica: C. jap?nica is univoltine and is distributed throughout Japan except Okinawa and eastern Hokkaido. Its mode of life does not differ essentially from that of most its congeners (Sakagami and Maeta, 1977). Females excavate nests in pithy cores of twigs or stems, prepare the brood cells serially, separated by pith partitions, and practice mass provisioning for immatures. The life cycle can be divided in seven periods or phases, Pi-P7 (Table 9). analysis of natural mfn: In Morioka (39?45'N), where this study was carried out, nests of C jap?nica are dug mainly in dry stems of Rubus, Kenia and other pithy plants growing in semi-shaded wood margins. From 1974 to 1980 many dry stems of Rubus palmatus var. coptophyllus were placed vertically in natural nest sites in the spring. The occupied nests were collected from late June to late July (Table 9, from the end of period P2 to P3), either in the evening or on cloudy/ rainy days when the adult inhabitants were presumably all in their nests. As C. 1 Institute of Low Temperature Science, Hokkaido University, Sapporo 060, Japan. 2 Laboratory of Insect Management, Faculty of Agriculture, Shimane University, Matsue 690, Japan. Accepted for publication 26 January 1984. This content downloaded from 207.46.13.28 on Wed, 31 Aug 2016 04:18:02 UTC All use subject to http://about.jstor.org/terms 640 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY Fig. 1. (left) Nest room for rearing of Ceratina jap?nica, a wooden stand supporting artificial nests at the center and sweet clover as food source at the right, (right) Wooden stand for artificial glass-tube nests covered with black paper. jap?nica often reuses nest burrows constructed in the prior year, a number of occupied stems were always left at the sites and examined the next year. The following features are used in this study. I. Number of immatures of each stage and of adult females. II. Size of each female. Divided in six classes, S0-S5, by head width (Table 4). III. Age of each female. In this species more than 50% of the mothers born in the previous year pass through their second winter and many of them successfully produce a second brood (Sakagami and Maeta, 1977). Based on wing and man dibular wear, females were divided into new (Aq, those emerged the previous autumn) and older (A,). IV. Spermathecal condition, inseminated (Fi) or not (F0). V. Ovarian condition. Using the criteria established by Kurihara et al. (1980), females were divided in three classes: Egg layers (02). Ovaries contain oocytes of either stage IV (completely chorionated egg) or III (full grown oocyte before completion of chorionation), or if not, III' (oocyte degenerated from III). Previous egg layers (Ox). Ovaries do not contain IV, III, III' oocytes, but indicate previous laying activity by the enlarged calyces of some ovarioles which may contain corpora lutea. Non-egg layers (O0). Ovaries are threadlike, containing only stage I, II (previtellogenic or early vitellogenic) or F, II' (degenerated from I or II) oocytes. The symbol O' designates the females parasitized by Zodion vsevoldi Zimina. Distinction of O, from O0 was difficult in some late nests, the females of which exhibited advanced ovarian degeneration. In such cases the number and stages of the immatures and ovarian conditions of cohabitants were taken into consideration. In difficult cases precedence in assigning the ovarian class was given to the occurrence of previous oviposition. artificial induction of mfn: In early spring, adult males and females (both Ai and Aq) in the P7 stage (Table 9) were liberated in several 3 m (1) x 1.5 m (w) x 2 m (h) vinyl roofed rooms set in a greenhouse (Fig. 1). Glass tubes 50 cm long and 5 mm in inner diameter were used for nest substrates. The pithy cores of Kenia jap?nica of appropriate diameter were inserted in the tubes, each core being furrowed along one side to form a straight guide along which the bee burrows. About 20-30 tubes were supported on a wooden stand (Fig. 1). Each tube was kept semihorizontal, 5 cm from its neighbors with the entrance directed about 8? This content downloaded from 207.46.13.28 on Wed, 31 Aug 2016 04:18:02 UTC All use subject to http://about.jstor.org/terms VOLUME 57, NUMBER 4 641 Table 1. Number of multifemale, solitary and orphan nests examined (respectively Nm, Ns, N0) and (in parentheses) their percentage ratios. %Nm Nest type Year Multifemale Solitary Orphan Total (Nm + N,)~l Newly built 1974-75* 2(1.1) 153(84.1) 27(14.8) 182 1.2 1976 0(0) 18(69.2) 8(30.8) 26