Exploring the density-dependent structure of blowfly populations by nonparametric additive modeling

Abundances of 12 laboratory populations of the greenbottle blowfly (Lucilia sericata) were recorded every two days for 776 d, with separate counts for larvae, pupae, and adults. Half of the populations were exposed to sublethal dosages of the toxic compound cadmium acetate; the remaining populations were considered controls. Initial density was low for half of the populations in each group, and high for the other half. In all populations, the adult abundance underwent sustained fluctuations. However, cadmium-exposed populations had smaller mean larval and adult densities, and fluctuations in adult abundance were less regular than for controls. Data from the first and the second half of the experimental period were analyzed separately in order to assess the effects of possible long-term changes in the dynamics on the estimates. Nonparametric (generalized) additive modeling (GAM) was used to investigate time series dynamics and, in particular, to explore the density-independent components and the structure of the density-dependent components of the system. Overall, cadmium populations had larger larva-to-adult survival rate and smaller adult survival rate than control populations, and for the second half of the experimental period the reproductive rate was smaller for cadmium populations than for control populations. Estimation of the density-dependent components suggested that survival from larva to adult depended nonlinearly on larval density and that increased larval density had a positive effect on larval survival at low densities. Furthermore, cadmium generally decreased vital rates. However, the analysis suggested that most of the observed differences in dynamical behavior between control and cadmium populations are related to differences in the density-independent components of the demographic rates, rather than differences in the density-dependent structure.