Search

The differentiation of workers into morphological subcastes (e.g., soldiers) represents an important evolutionary transition and is thought to improve division of labor in social insects. Soldiers occur in many ant and termite species, where they make up a small proportion of the workforce. A common assumption of worker caste evolution is that soldiers are behavioral specialists. Here, we report the first test of the "rare specialist" hypothesis in a eusocial bee. Colonies of the stingless bee Tetragonisca angustula are defended by a small group of morphologically differentiated soldiers. Contrary to the rare specialist hypothesis, we found that soldiers worked more (+34%-41%) and performed a greater variety of tasks (+23%-34%) than other workers, particularly early in life. Our results suggest a "rare elite" function of soldiers in T. angustula, that is, that they perform a disproportionately large amount of the work. Division of labor was based on a combination of temporal and physical castes, but soldiers transitioned faster from one task to the next. We discuss why the rare specialist assumption might not hold in species with a moderate degree of worker differentiation.

Diploid males have long been considered a curiosity contradictory to the haplo-diploid mode of sex determination in the Hymenoptera. In Apis mellifera, 'false' diploid male larvae are eliminated by worker cannibalism immediately after hatching. A 'cannibalism substance' produced by diploid drone larvae to induce worker-assisted suicide has been hypothesized, but it has never been detected. Diploid drones are only removed some hours after hatching. Older larvae are evidently not regarded as 'false males' and instead are regularly nursed by the brood-attending worker bees. As the pheromonal cues presumably are located on the surface of newly hatched bee larvae, we extracted the cuticular secretions and analyzed their chemical composition by gas chromatograph-mass spectrometry (GC-MS) analyses. Larvae were sexed and then reared in vitro for up to three days. The GC-MS pattern that was obtained, with alkanes as the major compounds, was compared between diploid and haploid drone larvae. We also examined some physical parameters of adult drones. There was no difference between diploid and haploid males in their weight at the day of emergence. The diploid adult drones had fewer wing hooks and smaller testes. The sperm DNA content was 0.30 and 0.15 pg per nucleus, giving an exact 2:1 ratio for the gametocytes of diploid and haploid drones, respectively. Vitellogenin was found in the hemolymph of both types of imaginal drones at 5 to 6 days, with a significantly lower titer in the diploids.

The ectoparasitic bee mite, Varroa destructor, is highly adapted to its natural and adopted honey bee hosts, Apis cerana and Apis mellifera. Adult females perforate the integument of bee pupae in such a way that they and their progeny can feed. We examined the wounds that founder females made, and usually found one, and rarely up to three, integumental wounds on pupae of A. mellifera multiply infested by V. destructor. The punctures were mainly on the 2nd abdominal sternite of the host. These perforations are used repeatedly as feeding sites by these hemolymph-sucking mites and by their progeny. The diameter of the wounds increased during pupal development. In brood cells containing 4-5 invading female mites and their progeny, healing of the wound is delayed, normally occurring just before the imaginal moult of the bee pupa. These wounds are subject to microbial infections, and they are relevant to the evolution of behavioral traits in these parasitic mites and their relations to host bees.

We developed a bioassay to measure the flying power of drone, in order to determine which drones could reach a drone congregation area. A wind tunnel was used to test unparasitized drones and drones slightly parasitized by one or two mites during pupal development, and counts were made of the number of spermatozoa that they produced. Drones parasitized with one mite flew as long as control drones (x= 6'55" and 6'48", respectively, P = 0.512); however, those that had been infested by two mites flew significantly less (x= 2'16", P<0.001). There was a significant positive correlation (P<0.01) between flight duration and the number of spermatozoa per drone in control group (r = 0.53), and in both the one mite (r = 0.43) and two mite (r = 0.54) groups. Drones infested during development with one or two mites produced 24 and 45% fewer sperm, respectively.

In the honey bee vitellogenin is present not only in the hemolymph of fertile queens, but also of infertile nurse workers in colonies. Drones lack this female-specific protein.Vitellogenin was also found in gynandromorphs, especially in those with an anteriorposterior distribution of the male/female mosaics. The occurrence of vitellogenin in gynandromorphs is age-dependent and socially conditioned as in worker bees. The female-specific regulation of vitellogenin synthesis is probably based on sex-specific properties of the differentiated fat body.

Two systems of steady extra-cellular currents were found along the surface of the telotrophicDysdercus ovarioles by means of a vibrating probe. The first covers the subgerminal tropharium and all the previtellogenic follicles. The current leaves the 3 or 4 small follicles of early euplasmic growth stages laterally and enters the syncytial tropharium. We presume that a similar intracellular current flows between the trophoplasm and the ooplasm which are interconnected by narrow nurse strands. Preliminary intracellular measurements indicate a potential gradient within this continuous cytoplasm, the ooplasm being electropositive to that of the tropharium. This current system fits into a model of polarized intracytoplasmic transport by electrophoresis. It is possible to explain the well known directed and selective flow of RNA from the tropharium via the nurse strands into the oocytes by means of such a model. The second current system occurs around every one of the 2 to 8 vitellogenic follicles. The pattern is completely different from that described for the first system. In the vitellogenic stages the current enters the follicle laterally all along the now much extended surface. It is balanced by a strong peak current which leaves the interfollicular region. As data on intracellular currents are not yet avialable, it is only a matter of speculation whether the circuit is closed through the ooplasm or only by a tangential loop through the follicle epithelium. The possible significance of this second current system for vitellogenin accumulation and uptake by the vitellogenic oocytes is also uncertain as yet.

1. The soluble proteins of liver, blood, and ovary of the zebrafish,Brachydanio rerio, were separated by electrophoresis on cellulose acetate strips (cellogel) and polyacrylamide gels (6%). The ontogenesis of specific protein spectra was studied in females of 1, 4 and 6 months and compared with adult males. In females the development of distinct protein patterns correlated with the first vitellogenic cycle could be observed. 2. In the course of the incorporation of the yolk system (known from previous cytological investigations) 2 protein fractions identified as yolk materials appear in the ovary. 3. The first of these fractions is present already in ovary homogenates from females of 4 months. It consists of a glycoprotein with a small electronegative charge at pH 9. As this fraction is found only in the ovary, it represents endogenous yolk material. 4. In the 6 months old adult females an additional band appears in the ovary. This becomes the dominating fraction and was identified as an electronegatively charged lipoprotein. Fractions with corresponding properties are also found to a slight extent in the serum and can be extracted from the liver. That material therefore represents exogenous yolk. Probably the liver is the site of vitellogenin synthesis. 5. The endo- and the exogenous yolk substances have the same electrophoretic mobility in polyacrylamide gel. 6. The endogenous yolk is certainly, the exogenous probably, a female specific protein.

The soluble proteins of the haemolymph, ovary and eggs of the honey bee were separated by electrophoresis on cellulose acetate foils and on polyacrylamide gels. The predominant haemolymph fraction (60-80%) is a female-specific protein. Corresponding fractions are found in the ovary and in uncleaved eggs. According to the results, it is highly probable that these fractions represent the yolk material or vitellogenin. InApis, this vitellogenin is represented by only one protein band which carries no carbohydrate components.The haemolymph titre of vitellogenic protein depends on the physiological state of the female bees. In the worker bees, haemolymph yolk proteins are detectable only during the nurse age, whereas in queens they can be found throughout the year. It may be noted that in the mated and egg producing queens the concentration of vitellogenin in the haemolymph is lower than in virgin queens and in adult non-laying queens.The rate ofin vivo protein synthesis was determined by injecting a mixture of 14 C-amino acids and measuring the radioactivity in pherogram bands using a methane flow counter. In the actively laying queen the rate of vitellogenin synthesis is 3 times higher than that of the non-sex-specific blood proteins. Approximately 85% of the labeled proteins which are released from the fat body are yolk material. In contrast to other insects, non-laying queens were also found to synthesize vitellogenin; however, the rate did not exceed that of the other serum proteins. The fate of yolk material in such females is unknown. In the ovary, synthesis of the euplasmic proteins occurs at a rate similar to that of the nonspecific haemolyph fractions. The appearance of radioactivity in the ovarian yolk is delayed compared with all other proteins. This indicates that these proteins are labeled byde novo synthesis, whereas the yolk is not built up within the ovary. This conclusion is further supported by the fact that the increase in specific radioactivity of ovarian yolk proteins occurs at a time when free 14 C-amino acids are no longer available in the haemolymph. Following the maximum intensity of labelling of yolk proteins the specific activity declines. From the data it can be estimated that the period of vitellogenesis for a single follicle is 2 days; 2 additional days of maturation, during which glycogen synthesis and chorion formation occur, are required before oviposition.

The effect of cold (4° C) on protein synthesis and glycogen content of intermediate egg follicles was studied. Under normal conditions (21°) these developmental stages were still free of carbohydrates. 1. As shown by autoradiography, in the cold the incorporation of amino acids was reduced to 1/4. No cold adaptation was observed within 2 weeks. The growth of the follicles was stopped. 2. A decrease in temperature brings above a storage of glycogen with the same frequency in the early and middle stages of oogenesis. 3. In permanent cold about 75% of the intermediate oocytes were storing glycogen. 4. After a short period of 2 cold days glycogen which had accumulated in young follicles was catabolized again. The oocyte development then continued normally. 5. After a long period of 10 cold days disarrangements occurred. Follicle development was retarded and became asynchronous or even atretic. 6. If eggs affected by a period of cold became ripe, they had normal developmental abilities. Several results indicate that the enzymes necessary for glycogen synthesis are already present but inactivated during intermediate stages of oogenesis. They can be activated by low temperature. Some possibilities of a hormonal or an intracellular regulation of precocious glycogen storage are discussed. The regulatory system normally releases glycogen synthesis only towards the end of oogenesis.

Fully grown oocytes ofMusca domestica contain large amounts of glycogen distributed in a characteristic pattern. Three cytoplasmic layers can be distinguished: 1. The periplasm which is free of carbohydrates and merely contains some lipid and protein yolk. 2. A zone of large glycogen clods. 3. Adjacent to this the central ooplasm where numerous lipid droplets and protein yolk spheres are found beside medium size glycogen clods. The glycogen areas are not surrounded by membranes, in contrast to the other yolk inclusions. Some possible interpretations of this ooplasmic pattern, which is already established during oogenesis, are discussed.