The sibling species Anopheles gambiae s.s. and Anopheles coluzzii are the most important vectors of human malaria in sub-Saharan Africa. They are thought to be undergoing speciation with gene flow with rare viable hybrids but are reproductively isolated by assortative mating and ecological divergence. As such important vectors, they are the focus of novel control strategies based on mosquito releases. One of the known drivers of their ecological speciation is their divergent larval ecological adaptation that is possibly linked to rice domestication in Africa. The potential impact of such divergence has never been integrated into mosquito rearing to improve alternative vector control methods such as Sterile Insect Technique (SIT) and Genetically Modified Mosquitoes (GMM), that are needed to accelerate the progress towards malaria elimination. These innovative vector control methods depend on mass rearing of millions of mosquitoes in a manner that is both efficient and economic, to produce mosquitoes that are of adequate quality, able to favourably compete with wild populations. In this thesis, we investigated the phenotypic plasticity of these sibling species to typical stressors in the rice field ecosystem with a focus on ammonia in their larval habitat. Experiments were conducted in small containers and in contrasted microcosms to test the direct effects of mineral water and increasing ammonia concentrations on larval development and to highlight divergent reaction norms between the sibling species. We also evaluated the use of zeolite to improve larval water quality management in An. gambiae s.l. insectary. To further understand the dynamics of the nitrogen cycle in larval rearing trays that led to larval mortality, we characterised their bacteria communities using 16S rRNA gene sequencing. Functional filters were applied to identify candidate bacteria species beneficial and detrimental to larval development and these were validated by qPCR. Our results suggest that genotype-by-environment interactions associated with rice domestication event in Africa are indeed an important driver of the eco-speciation between the sibling species. An. coluzzii was more tolerant to ammonia and rice-field like conditions supporting the idea that this may have driven its speciation from the ancestral An. gambiae s.s. We show that mineral water is beneficial for improved mosquito yield and phenotypic quality of adult mosquitoes in the insectary and this can be used to improve rearing protocols for these species. For the first time, we demonstrated that zeolite can be used to improve rearing results for An. gambiae s.l., providing a water conserving alternative for rearing mosquitoes for mass release programmes, especially in arid regions. Furthermore, the ensuing analyses of bacterial communities larval trays is also a novel endeavour which led to the identification of 1031 bacteria species and of several key species with various opportunities for further improvement of larval rearing towards mass release purposes and/or for novel direct vector control. In conclusion, we have made modest contributions towards the control of these malaria vectors and the fight to eliminate this multifaceted disease. It is therefore important that policy makers in malaria endemic countries ensure that policy reformations in irrigational agriculture and urbanization consider the impact of policy on these disease vectors that are of immense public health importance.