We advocate to study how adaptation of mosquito vectors to environmental modifications associated with global change impact their fitness and the life-history traits influencing vectorial capacity, in order to predict more accurately the epidemiological consequences of niche expansions and the spread of mosquito-borne pathogens. Such predictions are essential to adapt disease control programmes and avoid the emergence of vector-borne diseases. Toward this aim, we propose to adapt to the study of natural mosquito populations the multi-state/multi-event mark-recapture (MSMR) analytical approach, that has been instrumental to obtain unbiased field estimates of vertebrate populations demographic parameters. Individual capture histories will be obtained by 'marking' mosquitoes with genetic fingerprints, using environmental DNA collected non-invasively. We propose to verify whether the recent invasion of urban-polluted and coastal-brackish water habitats by populations of the Anopheles gambiae complex and Aedes aegypti, which are among the best vectors of malaria and dengue in the world, is adaptative, and to assess the cost of adaptation by comparing fitness trade-offs in reciprocally transplanted natural populations occurring in contrasting environments. To gain insights about the generality of the patterns observed, we propose to conduct these field and semi-field experiments using mosquito populations from four African countries and two environmental stressors, i.e. water pollution and salinity.