Abstract. 1. The work reported here investigated the incidence of dispersal by bagworm larvae Metisa plana (Walker) (Lepidoptera: Psychidae), a species with apterous females and winged males.2. A mark±recapture study conducted in a plantation of oil palm revealed that a significant proportion of larvae undertakes multiple episodes of ballooning, and suggested that ballooning represents a strategy for foraging both within and between hosts.3. Experiments conducted in a controlled cage environment indicated that increasing population density and defoliation of oil palm both promote dispersal by larvae.4. Neonatal larvae typically remained on the host where they emerged, indicating that pupation site of late-instar females influences the distribution of their progeny, and suggesting a high level of genetic relatedness in localised bagworm populations.5. Density-dependent dispersal may generate relatively uniform distributions of M. plana in plantations of oil palm, by simultaneously stabilising populations on heavily infested palms and redistributing larvae on lightly infested palms.6. The rate of ballooning was greater for female than for male larvae, possibly resulting from the sex-specific impact of population density on reproductive success or from late-instar females seeking suitable oviposition sites for their future progeny.

The observed mutualistic relationship between the black cocoa ant Dolichoderus thoracicus (Smith) and the mealybug Cataenococcus hispidus (Morrison) was examined. The importance of C. hispidus to D. thoracicus as a food source was investigated by giving D. thoracicus access to C. hispidus only, to C. hispidus and other food sources, and denying access to any obvious food sources. Dolichoderus thoracicus was seen to depend on C. hispidus alone as a source of food over an eight-week period of observation without showing ill effects. The role of D. thoracicus in spreading C. hispidus was studied in an experiment consisting of combinations of D. thoracicus and C. hispidus exclusion. It was shown that D. thoracicus was responsible for carrying C. hispidus across a ‘mealybug excluder’. Data on the frequency and duration of transport of C. hispidus by D. thoracicus were obtained by direct observation for a total of 90 h over 17 days. Both adults and nymphs of C. hispidus were carried by D. thoracicus with its mandibles in a brief and erratic manner, in the general direction of the trail. Such transportation is nevertheless considered to be important in view of the large number of individuals of D. thoracicus moving along a trail. The close mutualistic relationship between D. thoracicus and C. hispidus reiterates the necessity to manipulate both organisms for control of cocoa pests.

The influence of the black cocoa ant Dolichoderus thoracicus (Smith) on losses due to the mirid Helopeltis theivora Waterhouse, black pod disease (caused by Phytophthora palmivora (Peronosporales)) and mammalian pests (rats, squirrels and civet cats) was studied over a two-year period in two separate fields of cocoa. Both fields initially had high D. thoracicus activity. In each field, antscarce plots were created by treating these plots with insecticides; antabundant plots existed where the plots were left untreated. An abundance of D. thoracicus clearly had a negative effect on numbers of mirids, with respectively 380 and 2222% more nymphs and adults being recorded in the antscarce than in the ant-abundant plots over the two-year period. An abundance of D. thoracicus did not increase black pod incidence: on the contrary it reduced incidence of the disease in both fields. The number of rat-damaged pods in ant-scarce plots was significantly higher than in ant-abundant plots in one of the fields but no significant difference was detected in the other. Regarding number of pods lost to squirrels and civet cats, no significant difference between ant-scarce and ant-abundant plots was obtained for both fields; the losses to either of these mammalian pests were extremely low. The number of healthy ripe pods in ant-abundant plots was 40.4 and 32.1% higher than in ant-scarce plots in the two fields, these increases being ascribed to protection from H. theivora damage by D. thoracicus. IntroductionThe black cocoa ant Dolichoderus thoracicus (Smith) was first observed to have a negative effect on damage caused by Helopeltis bradyi Waterhouse (misidentified as H. antonii Signoret) and H. theivora Waterhouse (Heteroptera: Miridae) early this century in Indonesia (Giesberger, 1983). The effect was subsequently confirmed both

The unusual life history of bagworms, Metisa plana (Lepidoptera: Psychidae), with females completing all reproductive activity within a bag they construct as larvae, makes them ideal subjects to quantify intraspecific variations of reproductive success. This study investigated the relationship between body size and reproductive output of females, and tested the hypothesis that size-dependent variations of fecundity affect inter-generational variations of population density. Body size attained by female larvae at pupation (assessed by measuring the length of the pupal bag) was positively correlated with several parameters of reproductive output, including weight of pupae, weight and potential fecundity of calling female, as well as weight and realized fecundity of mated females. Our results further suggest that females allocate about two thirds of the resources they accumulate as larvae into egg production, and that a significant proportion of eggs are cannibalized by sibling neonates. In cage experiments, density of larvae in offspring generations on individual nursery palms was not affected by the size of females in parental generations. This result suggests that body size of females may not affect inter-generational variations of population density when potential resource depletion of host plants promotes a high incidence of dispersal among neonates.

No abstract