Abstract. Gaining a better understanding of parasite infection dynamics in the vector mosquito (Diptera: Culicidae) population is central to improving knowledge regarding the transmission, persistence and hence control of lymphatic filariasis. Here, we use data on mosquito feeding experiments collated from the published literature to examine the available evidence regarding the functional form of the first component of this parasite±vector relationship for Wuchereria bancrofti (Filarioidea: Onchocercidae) causing Bancroftian filariasis, i.e. the rate of microfilariae (mf) uptake from the blood of infected humans by the feeding mosquito vector. Using a simple logarithmic regression model for describing the observed relationships between the mean numbers of mf ingested per mosquito and parasite load in humans in each study, and a linear mixed-effects metaanalytical framework for synthesizing the observed regressions across studies, we show here for the first time clear evidence for the existence of densitydependence in this process for all the three major filariasis transmitting mosquito vectors. An important finding of this study is that this regulation of mf uptake also varies significantly between the vector genera, being weakest in Culex, comparatively stronger in Aedes and most severe and occurring at significantly lower human mf loads in Anopheles mosquitoes. The analysis of the corresponding mf uptake prevalence data has further highlighted how density-dependence in mf uptake may influence the observed distributions of mf in vector populations. These results show that whereas strong regulation of mf uptake, especially when it leads to saturation in uptake at low human parasite intensities, can lead to static distributions of mf per mosquito with host parasite intensity, a weaker regulation of mf ingestion can give rise to changes in both mean mf loads and in the frequency distribution of parasites/mosquito with increasing human parasite intensity. These findings highlight the importance of considering local vector infection dynamics when attempting to predict the impacts of community-based filariasis control. They also emphasize the value of developing and applying robust meta-analytic methods for estimating functional relationships regarding parasitic infection from population ecological data.

The principles of meta-analysis developed in a previous study were extended to investigate the process of Wuchereria bancrofti (Cobbold) (Filarioidea: Onchocercidae) infection in mosquito (Diptera: Culicidae) hosts, focusing specifically on the functional forms and strength of density dependence in the development of ingested microfilariae (mf) to infective (third instar) larvae (L3). Mathematical models describing observed mf-L3 functional responses for each of the major three parasite-transmitting vector genera, Aedes, Culex and Anopheles mosquitoes, were fitted to paired mf-L3 data collated from all available studies in the published literature. Model parameters were estimated and compared by deriving and applying a data synthetic framework, based on applying a non-linear weighted regression model for fitting mathematical models to multistudy data. The results confirm previous findings of the existence of significant between-genera differences in the mf-L3 development relationship, particularly with regard to the occurrence of limitation in Culex mosquitoes and facilitation in Aedes and Anopheles mosquitoes. New and unexpected findings regarding L3 development from ingested mf were discovered as follows: (1) for Culex, overcompensation in L3 development at higher intensities of mf (or a peaked mf-L3 functional response) was detected; (2) for Aedes mosquitoes, facilitation (with an apparent asymptotic constraint on L3 development at high mf densities) was shown to be the major process governing L3 development, and (3) for Anopheles, a stronger facilitation type of response with no apparent saturation in L3 development appears to govern L3 output from ingested mf. These results yield major new insights regarding filarial vector infection dynamics and their potential impacts on parasite control, and demonstrate the efficacy of employing a data synthetic approach to reveal and estimate parasitic infection processes in host populations.