Search

Experiments with sex allocation in parasitic wasps offer excellent opportunities for testing how the way in which organisms process information about their environment influences behaviour. If mating takes place in temporary patches, where only a small number of females produce offspring, then sex allocation theory predicts a female-biased sex ratio. When females lay different numbers of offspring in a patch, females that produce relatively fewer offspring should lay a less female-biased, or even male-biased, sex ratio. Recent theoretical models have predicted that the exact form of this relationship depends upon whether females know only their own clutch size (self knowledge) or also the clutch sizes laid by the other females on the patch (complete knowledge). We tested the predictions of these models by examining sex allocation when two females of the parasitoid wasp Nasonia vitripennis oviposited simultaneously on a patch. The offspring sex ratio (proportion of males) produced by a female was: (1) negatively correlated with the number of offspring that she laid; and (2) positively correlated with the body size of the other female on the patch. Larger females matured more eggs and laid more offspring in the experimental patch. This suggests that, as predicted by the complete knowledge model, the offspring sex ratio laid by a female became more female biased as she laid a greater proportion of the total offspring laid on the patch. Furthermore, females use the body size of other females to assess the clutch sizes that these will lay. Copyright 1998 The Association for the Study of Animal Behaviour

In haplodiploid species, the presence of unmated (virgin) females that can produce only haploid male offspring may have several important effects on a range of phenomena such as reproductive strategies, the transmission of parasitic chromosomes and the evolution of eusociality. The strength of these effects will depend upon the prevalence of virgin females. Two theories make conflicting predictions concerning the importance of factors that should be associated with increased levels of virginity, emphasizing either the degree of local mate competition or brood size. In this paper, a model is presented which predicts that, under conditions of local mate competition, the prevalence of virginity should be negatively correlated with the average number of offspring developing in a patch. The different predictions were then tested using data from 53 species of fig wasps representing 15 genera from four continents. Across species, the estimated prevalence of virginity was significantly inversely related to brood size, but showed no correlation with sex ratio (an index of local mate competition), supporting the predictions of our model. Qualitatively similar results were found when a formal comparative analysis was carried out using a morphologically and molecularly based phylogeny.

Models of insect-pathogen interactions in highly seasonal environments are developed. The models apply to insects such as many temperate forest pests that have a single generation per year and which are susceptible to viral disease only during their larval period. The disease kills the hosts after a fixed time period when infectious pathogen particles are released into the environment. Depending on the time taken to kill the host, one to many cycles of pathogen replication may occur during the portion of the year when susceptible hosts are present. A baseline model with linear disease transmission is always unstable although a stable equilibrium can be achieved if there is sufficient density dependence in the transmission process. Persistent, long-period cycles are virtually never observed. The release of pathogen particles prior to host death contributes towards stability although it does not result in limit cycles. Long-period cycles were found in two other extensions of the baseline model, one in which some hosts carry a sublethal infection which is transmitted to their offspring; and a second which includes a reservoir where pathogen particles are relatively long lived although unable to cause new infections. The relationship between this work and previous host-pathogen and host-parasitoid models is discussed.